Add files using upload-large-folder tool
Browse filesThis view is limited to 50 files because it contains too many changes. See raw diff
- ood-vifactcheck-bamibert-source-source-00-s43-20260829T002314Z-3858bb98/trainer/checkpoint-1456/special_tokens_map.json +51 -0
- ood-vifactcheck-bamibert-source-source-00-s43-20260829T002314Z-3858bb98/trainer/checkpoint-1456/tokenizer.json +0 -0
- ood-vifactcheck-bamibert-source-source-00-s43-20260829T002314Z-3858bb98/trainer/checkpoint-1456/tokenizer_config.json +55 -0
- ood-vifactcheck-bamibert-source-source-00-s43-20260829T002314Z-3858bb98/trainer/checkpoint-1456/trainer_state.json +537 -0
- ood-vifactcheck-bamibert-source-source-00-s43-20260829T002314Z-3858bb98/trainer/checkpoint-2002/special_tokens_map.json +51 -0
- ood-vifactcheck-bamibert-source-source-00-s43-20260829T002314Z-3858bb98/trainer/checkpoint-2002/tokenizer.json +0 -0
- ood-vifactcheck-bamibert-source-source-00-s43-20260829T002314Z-3858bb98/trainer/checkpoint-2002/tokenizer_config.json +55 -0
- ood-vifactcheck-bamibert-source-source-00-s43-20260829T002314Z-3858bb98/trainer/checkpoint-2002/trainer_state.json +724 -0
- ood-vifactcheck-bamibert-source-source-01-s43-20260829T003332Z-3858bb98/trainer/checkpoint-1368/special_tokens_map.json +51 -0
- ood-vifactcheck-bamibert-source-source-01-s43-20260829T003332Z-3858bb98/trainer/checkpoint-1368/tokenizer.json +0 -0
- ood-vifactcheck-bamibert-source-source-01-s43-20260829T003332Z-3858bb98/trainer/checkpoint-1368/tokenizer_config.json +55 -0
- ood-vifactcheck-bamibert-source-source-01-s43-20260829T003332Z-3858bb98/trainer/checkpoint-1368/trainer_state.json +509 -0
- ood-vifactcheck-bamibert-source-source-01-s43-20260829T003332Z-3858bb98/trainer/checkpoint-1881/special_tokens_map.json +51 -0
- ood-vifactcheck-bamibert-source-source-01-s43-20260829T003332Z-3858bb98/trainer/checkpoint-1881/tokenizer.json +0 -0
- ood-vifactcheck-bamibert-source-source-01-s43-20260829T003332Z-3858bb98/trainer/checkpoint-1881/tokenizer_config.json +55 -0
- ood-vifactcheck-bamibert-source-source-01-s43-20260829T003332Z-3858bb98/trainer/checkpoint-1881/trainer_state.json +689 -0
- ood-vifactcheck-bamibert-source-source-02-s43-20260829T004338Z-3858bb98/trainer/checkpoint-2715/special_tokens_map.json +51 -0
- ood-vifactcheck-bamibert-source-source-02-s43-20260829T004338Z-3858bb98/trainer/checkpoint-2715/tokenizer.json +0 -0
- ood-vifactcheck-bamibert-source-source-02-s43-20260829T004338Z-3858bb98/trainer/checkpoint-2715/tokenizer_config.json +55 -0
- ood-vifactcheck-bamibert-source-source-02-s43-20260829T004338Z-3858bb98/trainer/checkpoint-2715/trainer_state.json +964 -0
- ood-vifactcheck-bamibert-source-source-02-s43-20260829T004338Z-3858bb98/trainer/checkpoint-3258/special_tokens_map.json +51 -0
- ood-vifactcheck-bamibert-source-source-02-s43-20260829T004338Z-3858bb98/trainer/checkpoint-3258/tokenizer.json +0 -0
- ood-vifactcheck-bamibert-source-source-02-s43-20260829T004338Z-3858bb98/trainer/checkpoint-3258/tokenizer_config.json +55 -0
- ood-vifactcheck-bamibert-source-source-02-s43-20260829T004338Z-3858bb98/trainer/checkpoint-3258/trainer_state.json +1151 -0
- ood-vifactcheck-bamibert-source-source-03-s42-20260828T232240Z-6235c159/trainer/checkpoint-1464/special_tokens_map.json +51 -0
- ood-vifactcheck-bamibert-source-source-03-s42-20260828T232240Z-6235c159/trainer/checkpoint-1464/tokenizer.json +0 -0
- ood-vifactcheck-bamibert-source-source-03-s42-20260828T232240Z-6235c159/trainer/checkpoint-1464/trainer_state.json +537 -0
- ood-vifactcheck-bamibert-source-source-03-s43-20260829T010007Z-3858bb98/trainer/checkpoint-1647/special_tokens_map.json +51 -0
- ood-vifactcheck-bamibert-source-source-03-s43-20260829T010007Z-3858bb98/trainer/checkpoint-1647/tokenizer.json +0 -0
- ood-vifactcheck-bamibert-source-source-03-s43-20260829T010007Z-3858bb98/trainer/checkpoint-1647/tokenizer_config.json +55 -0
- ood-vifactcheck-bamibert-source-source-03-s43-20260829T010007Z-3858bb98/trainer/checkpoint-1647/trainer_state.json +597 -0
- ood-vifactcheck-bamibert-source-source-03-s43-20260829T010007Z-3858bb98/trainer/checkpoint-2196/special_tokens_map.json +51 -0
- ood-vifactcheck-bamibert-source-source-03-s43-20260829T010007Z-3858bb98/trainer/checkpoint-2196/tokenizer.json +0 -0
- ood-vifactcheck-bamibert-source-source-03-s43-20260829T010007Z-3858bb98/trainer/checkpoint-2196/tokenizer_config.json +55 -0
- ood-vifactcheck-bamibert-source-source-03-s43-20260829T010007Z-3858bb98/trainer/checkpoint-2196/trainer_state.json +784 -0
- ood-vifactcheck-bamibert-source-source-04-s42-20260828T233102Z-6235c159/trainer/checkpoint-1760/special_tokens_map.json +51 -0
- ood-vifactcheck-bamibert-source-source-04-s42-20260828T233102Z-6235c159/trainer/checkpoint-1760/tokenizer.json +0 -0
- ood-vifactcheck-bamibert-source-source-04-s42-20260828T233102Z-6235c159/trainer/checkpoint-1760/tokenizer_config.json +55 -0
- ood-vifactcheck-bamibert-source-source-04-s42-20260828T233102Z-6235c159/trainer/checkpoint-1760/trainer_state.json +643 -0
- ood-vifactcheck-bamibert-source-source-04-s42-20260828T233102Z-6235c159/trainer/checkpoint-2288/special_tokens_map.json +51 -0
- ood-vifactcheck-bamibert-source-source-04-s42-20260828T233102Z-6235c159/trainer/checkpoint-2288/tokenizer.json +0 -0
- ood-vifactcheck-bamibert-source-source-04-s42-20260828T233102Z-6235c159/trainer/checkpoint-2288/tokenizer_config.json +55 -0
- ood-vifactcheck-bamibert-source-source-04-s42-20260828T233102Z-6235c159/trainer/checkpoint-2288/trainer_state.json +823 -0
- ood-vifactcheck-bamibert-source-source-04-s43-20260829T011110Z-3858bb98/trainer/checkpoint-1408/special_tokens_map.json +51 -0
- ood-vifactcheck-bamibert-source-source-04-s43-20260829T011110Z-3858bb98/trainer/checkpoint-1408/tokenizer.json +0 -0
- ood-vifactcheck-bamibert-source-source-04-s43-20260829T011110Z-3858bb98/trainer/checkpoint-1408/tokenizer_config.json +55 -0
- ood-vifactcheck-bamibert-source-source-04-s43-20260829T011110Z-3858bb98/trainer/checkpoint-1408/trainer_state.json +523 -0
- ood-vifactcheck-bamibert-source-source-04-s43-20260829T011110Z-3858bb98/trainer/checkpoint-880/special_tokens_map.json +51 -0
- ood-vifactcheck-bamibert-source-source-04-s43-20260829T011110Z-3858bb98/trainer/checkpoint-880/tokenizer.json +0 -0
- ood-vifactcheck-bamibert-source-source-04-s43-20260829T011110Z-3858bb98/trainer/checkpoint-880/tokenizer_config.json +55 -0
ood-vifactcheck-bamibert-source-source-00-s43-20260829T002314Z-3858bb98/trainer/checkpoint-1456/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-source-source-00-s43-20260829T002314Z-3858bb98/trainer/checkpoint-1456/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-vifactcheck-bamibert-source-source-00-s43-20260829T002314Z-3858bb98/trainer/checkpoint-1456/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-source-source-00-s43-20260829T002314Z-3858bb98/trainer/checkpoint-1456/trainer_state.json
ADDED
|
@@ -0,0 +1,537 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1456,
|
| 3 |
+
"best_metric": 0.8094781108587298,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-source-source-00-s43-20260829T002314Z-3858bb98/trainer/checkpoint-1456",
|
| 5 |
+
"epoch": 8.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1456,
|
| 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": 7.192019939422607,
|
| 15 |
+
"learning_rate": 1.3186813186813187e-06,
|
| 16 |
+
"loss": 4.4325,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.27586206896551724,
|
| 21 |
+
"grad_norm": 97.57815551757812,
|
| 22 |
+
"learning_rate": 2.6923076923076923e-06,
|
| 23 |
+
"loss": 4.4086,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.41379310344827586,
|
| 28 |
+
"grad_norm": 7.4102373123168945,
|
| 29 |
+
"learning_rate": 4.065934065934066e-06,
|
| 30 |
+
"loss": 4.3891,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5517241379310345,
|
| 35 |
+
"grad_norm": 7.212603569030762,
|
| 36 |
+
"learning_rate": 5.43956043956044e-06,
|
| 37 |
+
"loss": 4.3915,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.6896551724137931,
|
| 42 |
+
"grad_norm": 9.142982482910156,
|
| 43 |
+
"learning_rate": 6.813186813186814e-06,
|
| 44 |
+
"loss": 4.3495,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.8275862068965517,
|
| 49 |
+
"grad_norm": 17.602365493774414,
|
| 50 |
+
"learning_rate": 8.186813186813188e-06,
|
| 51 |
+
"loss": 4.2514,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.9655172413793104,
|
| 56 |
+
"grad_norm": 24.169984817504883,
|
| 57 |
+
"learning_rate": 9.560439560439562e-06,
|
| 58 |
+
"loss": 4.1755,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 1.0,
|
| 63 |
+
"eval_accuracy": 0.4370015948963317,
|
| 64 |
+
"eval_loss": 1.0175344944000244,
|
| 65 |
+
"eval_macro_f1": 0.3458661760316877,
|
| 66 |
+
"eval_runtime": 1.1863,
|
| 67 |
+
"eval_samples_per_second": 528.549,
|
| 68 |
+
"eval_steps_per_second": 33.719,
|
| 69 |
+
"eval_weighted_f1": 0.348475318843276,
|
| 70 |
+
"step": 182
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.0993103448275863,
|
| 74 |
+
"grad_norm": 124.35462951660156,
|
| 75 |
+
"learning_rate": 1.0934065934065933e-05,
|
| 76 |
+
"loss": 3.803,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.237241379310345,
|
| 81 |
+
"grad_norm": 26.238130569458008,
|
| 82 |
+
"learning_rate": 1.230769230769231e-05,
|
| 83 |
+
"loss": 3.6348,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.3751724137931034,
|
| 88 |
+
"grad_norm": 39.70674133300781,
|
| 89 |
+
"learning_rate": 1.3681318681318683e-05,
|
| 90 |
+
"loss": 3.227,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.513103448275862,
|
| 95 |
+
"grad_norm": 86.24166107177734,
|
| 96 |
+
"learning_rate": 1.5054945054945056e-05,
|
| 97 |
+
"loss": 2.923,
|
| 98 |
+
"step": 275
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.6510344827586207,
|
| 102 |
+
"grad_norm": 27.281124114990234,
|
| 103 |
+
"learning_rate": 1.642857142857143e-05,
|
| 104 |
+
"loss": 2.9769,
|
| 105 |
+
"step": 300
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.7889655172413792,
|
| 109 |
+
"grad_norm": 66.61781311035156,
|
| 110 |
+
"learning_rate": 1.78021978021978e-05,
|
| 111 |
+
"loss": 2.4583,
|
| 112 |
+
"step": 325
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 1.926896551724138,
|
| 116 |
+
"grad_norm": 27.42291259765625,
|
| 117 |
+
"learning_rate": 1.9175824175824178e-05,
|
| 118 |
+
"loss": 2.4619,
|
| 119 |
+
"step": 350
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 2.0,
|
| 123 |
+
"eval_accuracy": 0.7543859649122807,
|
| 124 |
+
"eval_loss": 0.6403518319129944,
|
| 125 |
+
"eval_macro_f1": 0.7546829917400034,
|
| 126 |
+
"eval_runtime": 1.1305,
|
| 127 |
+
"eval_samples_per_second": 554.603,
|
| 128 |
+
"eval_steps_per_second": 35.381,
|
| 129 |
+
"eval_weighted_f1": 0.7544949658558368,
|
| 130 |
+
"step": 364
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.0606896551724136,
|
| 134 |
+
"grad_norm": 63.8184814453125,
|
| 135 |
+
"learning_rate": 1.993894993894994e-05,
|
| 136 |
+
"loss": 1.8778,
|
| 137 |
+
"step": 375
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 2.1986206896551725,
|
| 141 |
+
"grad_norm": 39.625308990478516,
|
| 142 |
+
"learning_rate": 1.9786324786324788e-05,
|
| 143 |
+
"loss": 1.9096,
|
| 144 |
+
"step": 400
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.336551724137931,
|
| 148 |
+
"grad_norm": 26.303911209106445,
|
| 149 |
+
"learning_rate": 1.9633699633699634e-05,
|
| 150 |
+
"loss": 1.7458,
|
| 151 |
+
"step": 425
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.47448275862069,
|
| 155 |
+
"grad_norm": 33.64207458496094,
|
| 156 |
+
"learning_rate": 1.948107448107448e-05,
|
| 157 |
+
"loss": 2.1433,
|
| 158 |
+
"step": 450
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.6124137931034483,
|
| 162 |
+
"grad_norm": 57.375850677490234,
|
| 163 |
+
"learning_rate": 1.932844932844933e-05,
|
| 164 |
+
"loss": 1.8675,
|
| 165 |
+
"step": 475
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.750344827586207,
|
| 169 |
+
"grad_norm": 35.25626754760742,
|
| 170 |
+
"learning_rate": 1.9175824175824178e-05,
|
| 171 |
+
"loss": 2.0076,
|
| 172 |
+
"step": 500
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 2.8882758620689657,
|
| 176 |
+
"grad_norm": 45.43314743041992,
|
| 177 |
+
"learning_rate": 1.9023199023199025e-05,
|
| 178 |
+
"loss": 2.2281,
|
| 179 |
+
"step": 525
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 3.0,
|
| 183 |
+
"eval_accuracy": 0.7671451355661882,
|
| 184 |
+
"eval_loss": 0.6163336038589478,
|
| 185 |
+
"eval_macro_f1": 0.7671848338264532,
|
| 186 |
+
"eval_runtime": 1.1809,
|
| 187 |
+
"eval_samples_per_second": 530.958,
|
| 188 |
+
"eval_steps_per_second": 33.873,
|
| 189 |
+
"eval_weighted_f1": 0.7669353340058069,
|
| 190 |
+
"step": 546
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 3.0220689655172412,
|
| 194 |
+
"grad_norm": 47.18822479248047,
|
| 195 |
+
"learning_rate": 1.887057387057387e-05,
|
| 196 |
+
"loss": 1.761,
|
| 197 |
+
"step": 550
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 3.16,
|
| 201 |
+
"grad_norm": 62.35240936279297,
|
| 202 |
+
"learning_rate": 1.8717948717948718e-05,
|
| 203 |
+
"loss": 1.3959,
|
| 204 |
+
"step": 575
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.2979310344827586,
|
| 208 |
+
"grad_norm": 68.5652847290039,
|
| 209 |
+
"learning_rate": 1.8565323565323568e-05,
|
| 210 |
+
"loss": 1.3324,
|
| 211 |
+
"step": 600
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.435862068965517,
|
| 215 |
+
"grad_norm": 43.91178894042969,
|
| 216 |
+
"learning_rate": 1.8412698412698415e-05,
|
| 217 |
+
"loss": 1.2364,
|
| 218 |
+
"step": 625
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.573793103448276,
|
| 222 |
+
"grad_norm": 46.89796447753906,
|
| 223 |
+
"learning_rate": 1.826007326007326e-05,
|
| 224 |
+
"loss": 1.303,
|
| 225 |
+
"step": 650
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.7117241379310344,
|
| 229 |
+
"grad_norm": 20.92741584777832,
|
| 230 |
+
"learning_rate": 1.8107448107448108e-05,
|
| 231 |
+
"loss": 1.2794,
|
| 232 |
+
"step": 675
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.849655172413793,
|
| 236 |
+
"grad_norm": 42.822853088378906,
|
| 237 |
+
"learning_rate": 1.7954822954822958e-05,
|
| 238 |
+
"loss": 1.4243,
|
| 239 |
+
"step": 700
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 3.9875862068965517,
|
| 243 |
+
"grad_norm": 27.365585327148438,
|
| 244 |
+
"learning_rate": 1.78021978021978e-05,
|
| 245 |
+
"loss": 1.3363,
|
| 246 |
+
"step": 725
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 4.0,
|
| 250 |
+
"eval_accuracy": 0.8038277511961722,
|
| 251 |
+
"eval_loss": 0.5572702288627625,
|
| 252 |
+
"eval_macro_f1": 0.8029159048831623,
|
| 253 |
+
"eval_runtime": 1.1818,
|
| 254 |
+
"eval_samples_per_second": 530.566,
|
| 255 |
+
"eval_steps_per_second": 33.848,
|
| 256 |
+
"eval_weighted_f1": 0.8029969569936386,
|
| 257 |
+
"step": 728
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 4.121379310344827,
|
| 261 |
+
"grad_norm": 44.50846481323242,
|
| 262 |
+
"learning_rate": 1.764957264957265e-05,
|
| 263 |
+
"loss": 0.9056,
|
| 264 |
+
"step": 750
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.259310344827586,
|
| 268 |
+
"grad_norm": 104.5069808959961,
|
| 269 |
+
"learning_rate": 1.74969474969475e-05,
|
| 270 |
+
"loss": 0.9929,
|
| 271 |
+
"step": 775
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 4.397241379310345,
|
| 275 |
+
"grad_norm": 43.41699981689453,
|
| 276 |
+
"learning_rate": 1.7344322344322345e-05,
|
| 277 |
+
"loss": 0.9497,
|
| 278 |
+
"step": 800
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 4.535172413793103,
|
| 282 |
+
"grad_norm": 20.099977493286133,
|
| 283 |
+
"learning_rate": 1.7191697191697195e-05,
|
| 284 |
+
"loss": 0.9549,
|
| 285 |
+
"step": 825
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 4.673103448275862,
|
| 289 |
+
"grad_norm": 64.63874053955078,
|
| 290 |
+
"learning_rate": 1.703907203907204e-05,
|
| 291 |
+
"loss": 0.6489,
|
| 292 |
+
"step": 850
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 4.811034482758621,
|
| 296 |
+
"grad_norm": 35.486934661865234,
|
| 297 |
+
"learning_rate": 1.688644688644689e-05,
|
| 298 |
+
"loss": 0.9993,
|
| 299 |
+
"step": 875
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 4.94896551724138,
|
| 303 |
+
"grad_norm": 77.00880432128906,
|
| 304 |
+
"learning_rate": 1.6733821733821735e-05,
|
| 305 |
+
"loss": 0.6723,
|
| 306 |
+
"step": 900
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 5.0,
|
| 310 |
+
"eval_accuracy": 0.79585326953748,
|
| 311 |
+
"eval_loss": 0.6997920870780945,
|
| 312 |
+
"eval_macro_f1": 0.795388234152806,
|
| 313 |
+
"eval_runtime": 1.0876,
|
| 314 |
+
"eval_samples_per_second": 576.487,
|
| 315 |
+
"eval_steps_per_second": 36.777,
|
| 316 |
+
"eval_weighted_f1": 0.7956787863273896,
|
| 317 |
+
"step": 910
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 5.082758620689655,
|
| 321 |
+
"grad_norm": 74.78981018066406,
|
| 322 |
+
"learning_rate": 1.6581196581196585e-05,
|
| 323 |
+
"loss": 0.6909,
|
| 324 |
+
"step": 925
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 5.220689655172414,
|
| 328 |
+
"grad_norm": 95.67037963867188,
|
| 329 |
+
"learning_rate": 1.642857142857143e-05,
|
| 330 |
+
"loss": 0.5206,
|
| 331 |
+
"step": 950
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.358620689655172,
|
| 335 |
+
"grad_norm": 41.395687103271484,
|
| 336 |
+
"learning_rate": 1.627594627594628e-05,
|
| 337 |
+
"loss": 0.7583,
|
| 338 |
+
"step": 975
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 5.496551724137931,
|
| 342 |
+
"grad_norm": 72.8214340209961,
|
| 343 |
+
"learning_rate": 1.6123321123321125e-05,
|
| 344 |
+
"loss": 0.6511,
|
| 345 |
+
"step": 1000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.63448275862069,
|
| 349 |
+
"grad_norm": 64.16326904296875,
|
| 350 |
+
"learning_rate": 1.5970695970695972e-05,
|
| 351 |
+
"loss": 0.738,
|
| 352 |
+
"step": 1025
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 5.772413793103448,
|
| 356 |
+
"grad_norm": 26.486474990844727,
|
| 357 |
+
"learning_rate": 1.581807081807082e-05,
|
| 358 |
+
"loss": 0.5873,
|
| 359 |
+
"step": 1050
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 5.910344827586207,
|
| 363 |
+
"grad_norm": 78.82721710205078,
|
| 364 |
+
"learning_rate": 1.5665445665445665e-05,
|
| 365 |
+
"loss": 0.9554,
|
| 366 |
+
"step": 1075
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 6.0,
|
| 370 |
+
"eval_accuracy": 0.7767145135566188,
|
| 371 |
+
"eval_loss": 0.8545607924461365,
|
| 372 |
+
"eval_macro_f1": 0.7788960246979176,
|
| 373 |
+
"eval_runtime": 1.1481,
|
| 374 |
+
"eval_samples_per_second": 546.13,
|
| 375 |
+
"eval_steps_per_second": 34.841,
|
| 376 |
+
"eval_weighted_f1": 0.7785816838117848,
|
| 377 |
+
"step": 1092
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 6.0441379310344825,
|
| 381 |
+
"grad_norm": 18.11637306213379,
|
| 382 |
+
"learning_rate": 1.5512820512820516e-05,
|
| 383 |
+
"loss": 0.5878,
|
| 384 |
+
"step": 1100
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 6.182068965517241,
|
| 388 |
+
"grad_norm": 119.37727355957031,
|
| 389 |
+
"learning_rate": 1.5360195360195362e-05,
|
| 390 |
+
"loss": 0.5217,
|
| 391 |
+
"step": 1125
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 6.32,
|
| 395 |
+
"grad_norm": 77.48580932617188,
|
| 396 |
+
"learning_rate": 1.5207570207570207e-05,
|
| 397 |
+
"loss": 0.4937,
|
| 398 |
+
"step": 1150
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 6.457931034482758,
|
| 402 |
+
"grad_norm": 37.662681579589844,
|
| 403 |
+
"learning_rate": 1.5054945054945056e-05,
|
| 404 |
+
"loss": 0.5351,
|
| 405 |
+
"step": 1175
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 6.595862068965517,
|
| 409 |
+
"grad_norm": 74.43467712402344,
|
| 410 |
+
"learning_rate": 1.4902319902319904e-05,
|
| 411 |
+
"loss": 0.5768,
|
| 412 |
+
"step": 1200
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 6.733793103448276,
|
| 416 |
+
"grad_norm": 91.94828796386719,
|
| 417 |
+
"learning_rate": 1.4749694749694752e-05,
|
| 418 |
+
"loss": 0.5142,
|
| 419 |
+
"step": 1225
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 6.871724137931034,
|
| 423 |
+
"grad_norm": 6.0705156326293945,
|
| 424 |
+
"learning_rate": 1.4597069597069597e-05,
|
| 425 |
+
"loss": 0.4501,
|
| 426 |
+
"step": 1250
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 7.0,
|
| 430 |
+
"eval_accuracy": 0.8086124401913876,
|
| 431 |
+
"eval_loss": 0.8989853858947754,
|
| 432 |
+
"eval_macro_f1": 0.8090229761543041,
|
| 433 |
+
"eval_runtime": 1.1308,
|
| 434 |
+
"eval_samples_per_second": 554.451,
|
| 435 |
+
"eval_steps_per_second": 35.372,
|
| 436 |
+
"eval_weighted_f1": 0.8095177359570919,
|
| 437 |
+
"step": 1274
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 7.0055172413793105,
|
| 441 |
+
"grad_norm": 2.4033329486846924,
|
| 442 |
+
"learning_rate": 1.4444444444444446e-05,
|
| 443 |
+
"loss": 0.3644,
|
| 444 |
+
"step": 1275
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 7.1434482758620685,
|
| 448 |
+
"grad_norm": 0.4781976640224457,
|
| 449 |
+
"learning_rate": 1.4291819291819293e-05,
|
| 450 |
+
"loss": 0.2932,
|
| 451 |
+
"step": 1300
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 7.281379310344827,
|
| 455 |
+
"grad_norm": 101.791748046875,
|
| 456 |
+
"learning_rate": 1.4139194139194141e-05,
|
| 457 |
+
"loss": 0.3883,
|
| 458 |
+
"step": 1325
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 7.419310344827586,
|
| 462 |
+
"grad_norm": 70.01971435546875,
|
| 463 |
+
"learning_rate": 1.3986568986568988e-05,
|
| 464 |
+
"loss": 0.5719,
|
| 465 |
+
"step": 1350
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 7.557241379310344,
|
| 469 |
+
"grad_norm": 0.4199492037296295,
|
| 470 |
+
"learning_rate": 1.3833943833943834e-05,
|
| 471 |
+
"loss": 0.3206,
|
| 472 |
+
"step": 1375
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 7.695172413793103,
|
| 476 |
+
"grad_norm": 5.778674602508545,
|
| 477 |
+
"learning_rate": 1.3681318681318683e-05,
|
| 478 |
+
"loss": 0.3862,
|
| 479 |
+
"step": 1400
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 7.833103448275862,
|
| 483 |
+
"grad_norm": 191.3223876953125,
|
| 484 |
+
"learning_rate": 1.3528693528693531e-05,
|
| 485 |
+
"loss": 0.4562,
|
| 486 |
+
"step": 1425
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 7.971034482758621,
|
| 490 |
+
"grad_norm": 0.13703469932079315,
|
| 491 |
+
"learning_rate": 1.3376068376068376e-05,
|
| 492 |
+
"loss": 0.2584,
|
| 493 |
+
"step": 1450
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 8.0,
|
| 497 |
+
"eval_accuracy": 0.8118022328548644,
|
| 498 |
+
"eval_loss": 1.0755796432495117,
|
| 499 |
+
"eval_macro_f1": 0.8094781108587298,
|
| 500 |
+
"eval_runtime": 1.2236,
|
| 501 |
+
"eval_samples_per_second": 512.425,
|
| 502 |
+
"eval_steps_per_second": 32.691,
|
| 503 |
+
"eval_weighted_f1": 0.8095397464114683,
|
| 504 |
+
"step": 1456
|
| 505 |
+
}
|
| 506 |
+
],
|
| 507 |
+
"logging_steps": 25,
|
| 508 |
+
"max_steps": 3640,
|
| 509 |
+
"num_input_tokens_seen": 0,
|
| 510 |
+
"num_train_epochs": 20,
|
| 511 |
+
"save_steps": 500,
|
| 512 |
+
"stateful_callbacks": {
|
| 513 |
+
"EarlyStoppingCallback": {
|
| 514 |
+
"args": {
|
| 515 |
+
"early_stopping_patience": 3,
|
| 516 |
+
"early_stopping_threshold": 0.0
|
| 517 |
+
},
|
| 518 |
+
"attributes": {
|
| 519 |
+
"early_stopping_patience_counter": 0
|
| 520 |
+
}
|
| 521 |
+
},
|
| 522 |
+
"TrainerControl": {
|
| 523 |
+
"args": {
|
| 524 |
+
"should_epoch_stop": false,
|
| 525 |
+
"should_evaluate": false,
|
| 526 |
+
"should_log": false,
|
| 527 |
+
"should_save": true,
|
| 528 |
+
"should_training_stop": false
|
| 529 |
+
},
|
| 530 |
+
"attributes": {}
|
| 531 |
+
}
|
| 532 |
+
},
|
| 533 |
+
"total_flos": 0.0,
|
| 534 |
+
"train_batch_size": 8,
|
| 535 |
+
"trial_name": null,
|
| 536 |
+
"trial_params": null
|
| 537 |
+
}
|
ood-vifactcheck-bamibert-source-source-00-s43-20260829T002314Z-3858bb98/trainer/checkpoint-2002/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-source-source-00-s43-20260829T002314Z-3858bb98/trainer/checkpoint-2002/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-vifactcheck-bamibert-source-source-00-s43-20260829T002314Z-3858bb98/trainer/checkpoint-2002/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-source-source-00-s43-20260829T002314Z-3858bb98/trainer/checkpoint-2002/trainer_state.json
ADDED
|
@@ -0,0 +1,724 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1456,
|
| 3 |
+
"best_metric": 0.8094781108587298,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-source-source-00-s43-20260829T002314Z-3858bb98/trainer/checkpoint-1456",
|
| 5 |
+
"epoch": 11.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2002,
|
| 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": 7.192019939422607,
|
| 15 |
+
"learning_rate": 1.3186813186813187e-06,
|
| 16 |
+
"loss": 4.4325,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.27586206896551724,
|
| 21 |
+
"grad_norm": 97.57815551757812,
|
| 22 |
+
"learning_rate": 2.6923076923076923e-06,
|
| 23 |
+
"loss": 4.4086,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.41379310344827586,
|
| 28 |
+
"grad_norm": 7.4102373123168945,
|
| 29 |
+
"learning_rate": 4.065934065934066e-06,
|
| 30 |
+
"loss": 4.3891,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5517241379310345,
|
| 35 |
+
"grad_norm": 7.212603569030762,
|
| 36 |
+
"learning_rate": 5.43956043956044e-06,
|
| 37 |
+
"loss": 4.3915,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.6896551724137931,
|
| 42 |
+
"grad_norm": 9.142982482910156,
|
| 43 |
+
"learning_rate": 6.813186813186814e-06,
|
| 44 |
+
"loss": 4.3495,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.8275862068965517,
|
| 49 |
+
"grad_norm": 17.602365493774414,
|
| 50 |
+
"learning_rate": 8.186813186813188e-06,
|
| 51 |
+
"loss": 4.2514,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.9655172413793104,
|
| 56 |
+
"grad_norm": 24.169984817504883,
|
| 57 |
+
"learning_rate": 9.560439560439562e-06,
|
| 58 |
+
"loss": 4.1755,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 1.0,
|
| 63 |
+
"eval_accuracy": 0.4370015948963317,
|
| 64 |
+
"eval_loss": 1.0175344944000244,
|
| 65 |
+
"eval_macro_f1": 0.3458661760316877,
|
| 66 |
+
"eval_runtime": 1.1863,
|
| 67 |
+
"eval_samples_per_second": 528.549,
|
| 68 |
+
"eval_steps_per_second": 33.719,
|
| 69 |
+
"eval_weighted_f1": 0.348475318843276,
|
| 70 |
+
"step": 182
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.0993103448275863,
|
| 74 |
+
"grad_norm": 124.35462951660156,
|
| 75 |
+
"learning_rate": 1.0934065934065933e-05,
|
| 76 |
+
"loss": 3.803,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.237241379310345,
|
| 81 |
+
"grad_norm": 26.238130569458008,
|
| 82 |
+
"learning_rate": 1.230769230769231e-05,
|
| 83 |
+
"loss": 3.6348,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.3751724137931034,
|
| 88 |
+
"grad_norm": 39.70674133300781,
|
| 89 |
+
"learning_rate": 1.3681318681318683e-05,
|
| 90 |
+
"loss": 3.227,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.513103448275862,
|
| 95 |
+
"grad_norm": 86.24166107177734,
|
| 96 |
+
"learning_rate": 1.5054945054945056e-05,
|
| 97 |
+
"loss": 2.923,
|
| 98 |
+
"step": 275
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.6510344827586207,
|
| 102 |
+
"grad_norm": 27.281124114990234,
|
| 103 |
+
"learning_rate": 1.642857142857143e-05,
|
| 104 |
+
"loss": 2.9769,
|
| 105 |
+
"step": 300
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.7889655172413792,
|
| 109 |
+
"grad_norm": 66.61781311035156,
|
| 110 |
+
"learning_rate": 1.78021978021978e-05,
|
| 111 |
+
"loss": 2.4583,
|
| 112 |
+
"step": 325
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 1.926896551724138,
|
| 116 |
+
"grad_norm": 27.42291259765625,
|
| 117 |
+
"learning_rate": 1.9175824175824178e-05,
|
| 118 |
+
"loss": 2.4619,
|
| 119 |
+
"step": 350
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 2.0,
|
| 123 |
+
"eval_accuracy": 0.7543859649122807,
|
| 124 |
+
"eval_loss": 0.6403518319129944,
|
| 125 |
+
"eval_macro_f1": 0.7546829917400034,
|
| 126 |
+
"eval_runtime": 1.1305,
|
| 127 |
+
"eval_samples_per_second": 554.603,
|
| 128 |
+
"eval_steps_per_second": 35.381,
|
| 129 |
+
"eval_weighted_f1": 0.7544949658558368,
|
| 130 |
+
"step": 364
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.0606896551724136,
|
| 134 |
+
"grad_norm": 63.8184814453125,
|
| 135 |
+
"learning_rate": 1.993894993894994e-05,
|
| 136 |
+
"loss": 1.8778,
|
| 137 |
+
"step": 375
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 2.1986206896551725,
|
| 141 |
+
"grad_norm": 39.625308990478516,
|
| 142 |
+
"learning_rate": 1.9786324786324788e-05,
|
| 143 |
+
"loss": 1.9096,
|
| 144 |
+
"step": 400
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.336551724137931,
|
| 148 |
+
"grad_norm": 26.303911209106445,
|
| 149 |
+
"learning_rate": 1.9633699633699634e-05,
|
| 150 |
+
"loss": 1.7458,
|
| 151 |
+
"step": 425
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.47448275862069,
|
| 155 |
+
"grad_norm": 33.64207458496094,
|
| 156 |
+
"learning_rate": 1.948107448107448e-05,
|
| 157 |
+
"loss": 2.1433,
|
| 158 |
+
"step": 450
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.6124137931034483,
|
| 162 |
+
"grad_norm": 57.375850677490234,
|
| 163 |
+
"learning_rate": 1.932844932844933e-05,
|
| 164 |
+
"loss": 1.8675,
|
| 165 |
+
"step": 475
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.750344827586207,
|
| 169 |
+
"grad_norm": 35.25626754760742,
|
| 170 |
+
"learning_rate": 1.9175824175824178e-05,
|
| 171 |
+
"loss": 2.0076,
|
| 172 |
+
"step": 500
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 2.8882758620689657,
|
| 176 |
+
"grad_norm": 45.43314743041992,
|
| 177 |
+
"learning_rate": 1.9023199023199025e-05,
|
| 178 |
+
"loss": 2.2281,
|
| 179 |
+
"step": 525
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 3.0,
|
| 183 |
+
"eval_accuracy": 0.7671451355661882,
|
| 184 |
+
"eval_loss": 0.6163336038589478,
|
| 185 |
+
"eval_macro_f1": 0.7671848338264532,
|
| 186 |
+
"eval_runtime": 1.1809,
|
| 187 |
+
"eval_samples_per_second": 530.958,
|
| 188 |
+
"eval_steps_per_second": 33.873,
|
| 189 |
+
"eval_weighted_f1": 0.7669353340058069,
|
| 190 |
+
"step": 546
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 3.0220689655172412,
|
| 194 |
+
"grad_norm": 47.18822479248047,
|
| 195 |
+
"learning_rate": 1.887057387057387e-05,
|
| 196 |
+
"loss": 1.761,
|
| 197 |
+
"step": 550
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 3.16,
|
| 201 |
+
"grad_norm": 62.35240936279297,
|
| 202 |
+
"learning_rate": 1.8717948717948718e-05,
|
| 203 |
+
"loss": 1.3959,
|
| 204 |
+
"step": 575
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.2979310344827586,
|
| 208 |
+
"grad_norm": 68.5652847290039,
|
| 209 |
+
"learning_rate": 1.8565323565323568e-05,
|
| 210 |
+
"loss": 1.3324,
|
| 211 |
+
"step": 600
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.435862068965517,
|
| 215 |
+
"grad_norm": 43.91178894042969,
|
| 216 |
+
"learning_rate": 1.8412698412698415e-05,
|
| 217 |
+
"loss": 1.2364,
|
| 218 |
+
"step": 625
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.573793103448276,
|
| 222 |
+
"grad_norm": 46.89796447753906,
|
| 223 |
+
"learning_rate": 1.826007326007326e-05,
|
| 224 |
+
"loss": 1.303,
|
| 225 |
+
"step": 650
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.7117241379310344,
|
| 229 |
+
"grad_norm": 20.92741584777832,
|
| 230 |
+
"learning_rate": 1.8107448107448108e-05,
|
| 231 |
+
"loss": 1.2794,
|
| 232 |
+
"step": 675
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.849655172413793,
|
| 236 |
+
"grad_norm": 42.822853088378906,
|
| 237 |
+
"learning_rate": 1.7954822954822958e-05,
|
| 238 |
+
"loss": 1.4243,
|
| 239 |
+
"step": 700
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 3.9875862068965517,
|
| 243 |
+
"grad_norm": 27.365585327148438,
|
| 244 |
+
"learning_rate": 1.78021978021978e-05,
|
| 245 |
+
"loss": 1.3363,
|
| 246 |
+
"step": 725
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 4.0,
|
| 250 |
+
"eval_accuracy": 0.8038277511961722,
|
| 251 |
+
"eval_loss": 0.5572702288627625,
|
| 252 |
+
"eval_macro_f1": 0.8029159048831623,
|
| 253 |
+
"eval_runtime": 1.1818,
|
| 254 |
+
"eval_samples_per_second": 530.566,
|
| 255 |
+
"eval_steps_per_second": 33.848,
|
| 256 |
+
"eval_weighted_f1": 0.8029969569936386,
|
| 257 |
+
"step": 728
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 4.121379310344827,
|
| 261 |
+
"grad_norm": 44.50846481323242,
|
| 262 |
+
"learning_rate": 1.764957264957265e-05,
|
| 263 |
+
"loss": 0.9056,
|
| 264 |
+
"step": 750
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.259310344827586,
|
| 268 |
+
"grad_norm": 104.5069808959961,
|
| 269 |
+
"learning_rate": 1.74969474969475e-05,
|
| 270 |
+
"loss": 0.9929,
|
| 271 |
+
"step": 775
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 4.397241379310345,
|
| 275 |
+
"grad_norm": 43.41699981689453,
|
| 276 |
+
"learning_rate": 1.7344322344322345e-05,
|
| 277 |
+
"loss": 0.9497,
|
| 278 |
+
"step": 800
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 4.535172413793103,
|
| 282 |
+
"grad_norm": 20.099977493286133,
|
| 283 |
+
"learning_rate": 1.7191697191697195e-05,
|
| 284 |
+
"loss": 0.9549,
|
| 285 |
+
"step": 825
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 4.673103448275862,
|
| 289 |
+
"grad_norm": 64.63874053955078,
|
| 290 |
+
"learning_rate": 1.703907203907204e-05,
|
| 291 |
+
"loss": 0.6489,
|
| 292 |
+
"step": 850
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 4.811034482758621,
|
| 296 |
+
"grad_norm": 35.486934661865234,
|
| 297 |
+
"learning_rate": 1.688644688644689e-05,
|
| 298 |
+
"loss": 0.9993,
|
| 299 |
+
"step": 875
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 4.94896551724138,
|
| 303 |
+
"grad_norm": 77.00880432128906,
|
| 304 |
+
"learning_rate": 1.6733821733821735e-05,
|
| 305 |
+
"loss": 0.6723,
|
| 306 |
+
"step": 900
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 5.0,
|
| 310 |
+
"eval_accuracy": 0.79585326953748,
|
| 311 |
+
"eval_loss": 0.6997920870780945,
|
| 312 |
+
"eval_macro_f1": 0.795388234152806,
|
| 313 |
+
"eval_runtime": 1.0876,
|
| 314 |
+
"eval_samples_per_second": 576.487,
|
| 315 |
+
"eval_steps_per_second": 36.777,
|
| 316 |
+
"eval_weighted_f1": 0.7956787863273896,
|
| 317 |
+
"step": 910
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 5.082758620689655,
|
| 321 |
+
"grad_norm": 74.78981018066406,
|
| 322 |
+
"learning_rate": 1.6581196581196585e-05,
|
| 323 |
+
"loss": 0.6909,
|
| 324 |
+
"step": 925
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 5.220689655172414,
|
| 328 |
+
"grad_norm": 95.67037963867188,
|
| 329 |
+
"learning_rate": 1.642857142857143e-05,
|
| 330 |
+
"loss": 0.5206,
|
| 331 |
+
"step": 950
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.358620689655172,
|
| 335 |
+
"grad_norm": 41.395687103271484,
|
| 336 |
+
"learning_rate": 1.627594627594628e-05,
|
| 337 |
+
"loss": 0.7583,
|
| 338 |
+
"step": 975
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 5.496551724137931,
|
| 342 |
+
"grad_norm": 72.8214340209961,
|
| 343 |
+
"learning_rate": 1.6123321123321125e-05,
|
| 344 |
+
"loss": 0.6511,
|
| 345 |
+
"step": 1000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.63448275862069,
|
| 349 |
+
"grad_norm": 64.16326904296875,
|
| 350 |
+
"learning_rate": 1.5970695970695972e-05,
|
| 351 |
+
"loss": 0.738,
|
| 352 |
+
"step": 1025
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 5.772413793103448,
|
| 356 |
+
"grad_norm": 26.486474990844727,
|
| 357 |
+
"learning_rate": 1.581807081807082e-05,
|
| 358 |
+
"loss": 0.5873,
|
| 359 |
+
"step": 1050
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 5.910344827586207,
|
| 363 |
+
"grad_norm": 78.82721710205078,
|
| 364 |
+
"learning_rate": 1.5665445665445665e-05,
|
| 365 |
+
"loss": 0.9554,
|
| 366 |
+
"step": 1075
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 6.0,
|
| 370 |
+
"eval_accuracy": 0.7767145135566188,
|
| 371 |
+
"eval_loss": 0.8545607924461365,
|
| 372 |
+
"eval_macro_f1": 0.7788960246979176,
|
| 373 |
+
"eval_runtime": 1.1481,
|
| 374 |
+
"eval_samples_per_second": 546.13,
|
| 375 |
+
"eval_steps_per_second": 34.841,
|
| 376 |
+
"eval_weighted_f1": 0.7785816838117848,
|
| 377 |
+
"step": 1092
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 6.0441379310344825,
|
| 381 |
+
"grad_norm": 18.11637306213379,
|
| 382 |
+
"learning_rate": 1.5512820512820516e-05,
|
| 383 |
+
"loss": 0.5878,
|
| 384 |
+
"step": 1100
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 6.182068965517241,
|
| 388 |
+
"grad_norm": 119.37727355957031,
|
| 389 |
+
"learning_rate": 1.5360195360195362e-05,
|
| 390 |
+
"loss": 0.5217,
|
| 391 |
+
"step": 1125
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 6.32,
|
| 395 |
+
"grad_norm": 77.48580932617188,
|
| 396 |
+
"learning_rate": 1.5207570207570207e-05,
|
| 397 |
+
"loss": 0.4937,
|
| 398 |
+
"step": 1150
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 6.457931034482758,
|
| 402 |
+
"grad_norm": 37.662681579589844,
|
| 403 |
+
"learning_rate": 1.5054945054945056e-05,
|
| 404 |
+
"loss": 0.5351,
|
| 405 |
+
"step": 1175
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 6.595862068965517,
|
| 409 |
+
"grad_norm": 74.43467712402344,
|
| 410 |
+
"learning_rate": 1.4902319902319904e-05,
|
| 411 |
+
"loss": 0.5768,
|
| 412 |
+
"step": 1200
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 6.733793103448276,
|
| 416 |
+
"grad_norm": 91.94828796386719,
|
| 417 |
+
"learning_rate": 1.4749694749694752e-05,
|
| 418 |
+
"loss": 0.5142,
|
| 419 |
+
"step": 1225
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 6.871724137931034,
|
| 423 |
+
"grad_norm": 6.0705156326293945,
|
| 424 |
+
"learning_rate": 1.4597069597069597e-05,
|
| 425 |
+
"loss": 0.4501,
|
| 426 |
+
"step": 1250
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 7.0,
|
| 430 |
+
"eval_accuracy": 0.8086124401913876,
|
| 431 |
+
"eval_loss": 0.8989853858947754,
|
| 432 |
+
"eval_macro_f1": 0.8090229761543041,
|
| 433 |
+
"eval_runtime": 1.1308,
|
| 434 |
+
"eval_samples_per_second": 554.451,
|
| 435 |
+
"eval_steps_per_second": 35.372,
|
| 436 |
+
"eval_weighted_f1": 0.8095177359570919,
|
| 437 |
+
"step": 1274
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 7.0055172413793105,
|
| 441 |
+
"grad_norm": 2.4033329486846924,
|
| 442 |
+
"learning_rate": 1.4444444444444446e-05,
|
| 443 |
+
"loss": 0.3644,
|
| 444 |
+
"step": 1275
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 7.1434482758620685,
|
| 448 |
+
"grad_norm": 0.4781976640224457,
|
| 449 |
+
"learning_rate": 1.4291819291819293e-05,
|
| 450 |
+
"loss": 0.2932,
|
| 451 |
+
"step": 1300
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 7.281379310344827,
|
| 455 |
+
"grad_norm": 101.791748046875,
|
| 456 |
+
"learning_rate": 1.4139194139194141e-05,
|
| 457 |
+
"loss": 0.3883,
|
| 458 |
+
"step": 1325
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 7.419310344827586,
|
| 462 |
+
"grad_norm": 70.01971435546875,
|
| 463 |
+
"learning_rate": 1.3986568986568988e-05,
|
| 464 |
+
"loss": 0.5719,
|
| 465 |
+
"step": 1350
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 7.557241379310344,
|
| 469 |
+
"grad_norm": 0.4199492037296295,
|
| 470 |
+
"learning_rate": 1.3833943833943834e-05,
|
| 471 |
+
"loss": 0.3206,
|
| 472 |
+
"step": 1375
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 7.695172413793103,
|
| 476 |
+
"grad_norm": 5.778674602508545,
|
| 477 |
+
"learning_rate": 1.3681318681318683e-05,
|
| 478 |
+
"loss": 0.3862,
|
| 479 |
+
"step": 1400
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 7.833103448275862,
|
| 483 |
+
"grad_norm": 191.3223876953125,
|
| 484 |
+
"learning_rate": 1.3528693528693531e-05,
|
| 485 |
+
"loss": 0.4562,
|
| 486 |
+
"step": 1425
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 7.971034482758621,
|
| 490 |
+
"grad_norm": 0.13703469932079315,
|
| 491 |
+
"learning_rate": 1.3376068376068376e-05,
|
| 492 |
+
"loss": 0.2584,
|
| 493 |
+
"step": 1450
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 8.0,
|
| 497 |
+
"eval_accuracy": 0.8118022328548644,
|
| 498 |
+
"eval_loss": 1.0755796432495117,
|
| 499 |
+
"eval_macro_f1": 0.8094781108587298,
|
| 500 |
+
"eval_runtime": 1.2236,
|
| 501 |
+
"eval_samples_per_second": 512.425,
|
| 502 |
+
"eval_steps_per_second": 32.691,
|
| 503 |
+
"eval_weighted_f1": 0.8095397464114683,
|
| 504 |
+
"step": 1456
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 8.104827586206897,
|
| 508 |
+
"grad_norm": 11.363759994506836,
|
| 509 |
+
"learning_rate": 1.3223443223443224e-05,
|
| 510 |
+
"loss": 0.5233,
|
| 511 |
+
"step": 1475
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 8.242758620689655,
|
| 515 |
+
"grad_norm": 131.52108764648438,
|
| 516 |
+
"learning_rate": 1.3070818070818073e-05,
|
| 517 |
+
"loss": 0.3418,
|
| 518 |
+
"step": 1500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 8.380689655172414,
|
| 522 |
+
"grad_norm": 1.150276780128479,
|
| 523 |
+
"learning_rate": 1.291819291819292e-05,
|
| 524 |
+
"loss": 0.3078,
|
| 525 |
+
"step": 1525
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 8.518620689655172,
|
| 529 |
+
"grad_norm": 3.4429514408111572,
|
| 530 |
+
"learning_rate": 1.2765567765567766e-05,
|
| 531 |
+
"loss": 0.207,
|
| 532 |
+
"step": 1550
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 8.65655172413793,
|
| 536 |
+
"grad_norm": 8.24793529510498,
|
| 537 |
+
"learning_rate": 1.2612942612942613e-05,
|
| 538 |
+
"loss": 0.3808,
|
| 539 |
+
"step": 1575
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 8.79448275862069,
|
| 543 |
+
"grad_norm": 155.26858520507812,
|
| 544 |
+
"learning_rate": 1.2460317460317461e-05,
|
| 545 |
+
"loss": 0.3269,
|
| 546 |
+
"step": 1600
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 8.932413793103448,
|
| 550 |
+
"grad_norm": 1.0432220697402954,
|
| 551 |
+
"learning_rate": 1.230769230769231e-05,
|
| 552 |
+
"loss": 0.2367,
|
| 553 |
+
"step": 1625
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 9.0,
|
| 557 |
+
"eval_accuracy": 0.8086124401913876,
|
| 558 |
+
"eval_loss": 1.2189340591430664,
|
| 559 |
+
"eval_macro_f1": 0.8041650260401795,
|
| 560 |
+
"eval_runtime": 1.1108,
|
| 561 |
+
"eval_samples_per_second": 564.439,
|
| 562 |
+
"eval_steps_per_second": 36.009,
|
| 563 |
+
"eval_weighted_f1": 0.8047265581214531,
|
| 564 |
+
"step": 1638
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 9.066206896551725,
|
| 568 |
+
"grad_norm": 1.0959495306015015,
|
| 569 |
+
"learning_rate": 1.2155067155067155e-05,
|
| 570 |
+
"loss": 0.3641,
|
| 571 |
+
"step": 1650
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 9.204137931034483,
|
| 575 |
+
"grad_norm": 146.79014587402344,
|
| 576 |
+
"learning_rate": 1.2002442002442003e-05,
|
| 577 |
+
"loss": 0.2692,
|
| 578 |
+
"step": 1675
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 9.34206896551724,
|
| 582 |
+
"grad_norm": 0.0887238085269928,
|
| 583 |
+
"learning_rate": 1.1849816849816852e-05,
|
| 584 |
+
"loss": 0.205,
|
| 585 |
+
"step": 1700
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 9.48,
|
| 589 |
+
"grad_norm": 3.827786684036255,
|
| 590 |
+
"learning_rate": 1.1697191697191698e-05,
|
| 591 |
+
"loss": 0.2491,
|
| 592 |
+
"step": 1725
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 9.617931034482758,
|
| 596 |
+
"grad_norm": 0.12241188436746597,
|
| 597 |
+
"learning_rate": 1.1544566544566545e-05,
|
| 598 |
+
"loss": 0.1202,
|
| 599 |
+
"step": 1750
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 9.755862068965516,
|
| 603 |
+
"grad_norm": 0.08224964886903763,
|
| 604 |
+
"learning_rate": 1.1391941391941392e-05,
|
| 605 |
+
"loss": 0.4072,
|
| 606 |
+
"step": 1775
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 9.893793103448276,
|
| 610 |
+
"grad_norm": 179.3701629638672,
|
| 611 |
+
"learning_rate": 1.123931623931624e-05,
|
| 612 |
+
"loss": 0.2504,
|
| 613 |
+
"step": 1800
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 10.0,
|
| 617 |
+
"eval_accuracy": 0.8070175438596491,
|
| 618 |
+
"eval_loss": 1.198479175567627,
|
| 619 |
+
"eval_macro_f1": 0.8046667715519195,
|
| 620 |
+
"eval_runtime": 1.1778,
|
| 621 |
+
"eval_samples_per_second": 532.339,
|
| 622 |
+
"eval_steps_per_second": 33.961,
|
| 623 |
+
"eval_weighted_f1": 0.8048729462227129,
|
| 624 |
+
"step": 1820
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 10.027586206896551,
|
| 628 |
+
"grad_norm": 0.17110122740268707,
|
| 629 |
+
"learning_rate": 1.1086691086691088e-05,
|
| 630 |
+
"loss": 0.1367,
|
| 631 |
+
"step": 1825
|
| 632 |
+
},
|
| 633 |
+
{
|
| 634 |
+
"epoch": 10.16551724137931,
|
| 635 |
+
"grad_norm": 0.06116608530282974,
|
| 636 |
+
"learning_rate": 1.0934065934065933e-05,
|
| 637 |
+
"loss": 0.1984,
|
| 638 |
+
"step": 1850
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 10.303448275862069,
|
| 642 |
+
"grad_norm": 0.06166241317987442,
|
| 643 |
+
"learning_rate": 1.0781440781440782e-05,
|
| 644 |
+
"loss": 0.1217,
|
| 645 |
+
"step": 1875
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 10.441379310344828,
|
| 649 |
+
"grad_norm": 0.23280438780784607,
|
| 650 |
+
"learning_rate": 1.062881562881563e-05,
|
| 651 |
+
"loss": 0.0853,
|
| 652 |
+
"step": 1900
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"epoch": 10.579310344827586,
|
| 656 |
+
"grad_norm": 0.11427852511405945,
|
| 657 |
+
"learning_rate": 1.0476190476190477e-05,
|
| 658 |
+
"loss": 0.1944,
|
| 659 |
+
"step": 1925
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 10.717241379310344,
|
| 663 |
+
"grad_norm": 0.07819736748933792,
|
| 664 |
+
"learning_rate": 1.0323565323565325e-05,
|
| 665 |
+
"loss": 0.3351,
|
| 666 |
+
"step": 1950
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 10.855172413793104,
|
| 670 |
+
"grad_norm": 0.19782736897468567,
|
| 671 |
+
"learning_rate": 1.0170940170940172e-05,
|
| 672 |
+
"loss": 0.1805,
|
| 673 |
+
"step": 1975
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 10.993103448275862,
|
| 677 |
+
"grad_norm": 1.2474019527435303,
|
| 678 |
+
"learning_rate": 1.0018315018315019e-05,
|
| 679 |
+
"loss": 0.1624,
|
| 680 |
+
"step": 2000
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 11.0,
|
| 684 |
+
"eval_accuracy": 0.8054226475279107,
|
| 685 |
+
"eval_loss": 1.2768216133117676,
|
| 686 |
+
"eval_macro_f1": 0.8055921985103244,
|
| 687 |
+
"eval_runtime": 1.1374,
|
| 688 |
+
"eval_samples_per_second": 551.269,
|
| 689 |
+
"eval_steps_per_second": 35.169,
|
| 690 |
+
"eval_weighted_f1": 0.8059105426902677,
|
| 691 |
+
"step": 2002
|
| 692 |
+
}
|
| 693 |
+
],
|
| 694 |
+
"logging_steps": 25,
|
| 695 |
+
"max_steps": 3640,
|
| 696 |
+
"num_input_tokens_seen": 0,
|
| 697 |
+
"num_train_epochs": 20,
|
| 698 |
+
"save_steps": 500,
|
| 699 |
+
"stateful_callbacks": {
|
| 700 |
+
"EarlyStoppingCallback": {
|
| 701 |
+
"args": {
|
| 702 |
+
"early_stopping_patience": 3,
|
| 703 |
+
"early_stopping_threshold": 0.0
|
| 704 |
+
},
|
| 705 |
+
"attributes": {
|
| 706 |
+
"early_stopping_patience_counter": 3
|
| 707 |
+
}
|
| 708 |
+
},
|
| 709 |
+
"TrainerControl": {
|
| 710 |
+
"args": {
|
| 711 |
+
"should_epoch_stop": false,
|
| 712 |
+
"should_evaluate": false,
|
| 713 |
+
"should_log": false,
|
| 714 |
+
"should_save": true,
|
| 715 |
+
"should_training_stop": true
|
| 716 |
+
},
|
| 717 |
+
"attributes": {}
|
| 718 |
+
}
|
| 719 |
+
},
|
| 720 |
+
"total_flos": 0.0,
|
| 721 |
+
"train_batch_size": 8,
|
| 722 |
+
"trial_name": null,
|
| 723 |
+
"trial_params": null
|
| 724 |
+
}
|
ood-vifactcheck-bamibert-source-source-01-s43-20260829T003332Z-3858bb98/trainer/checkpoint-1368/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-source-source-01-s43-20260829T003332Z-3858bb98/trainer/checkpoint-1368/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-vifactcheck-bamibert-source-source-01-s43-20260829T003332Z-3858bb98/trainer/checkpoint-1368/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-source-source-01-s43-20260829T003332Z-3858bb98/trainer/checkpoint-1368/trainer_state.json
ADDED
|
@@ -0,0 +1,509 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1368,
|
| 3 |
+
"best_metric": 0.824686944370819,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-source-source-01-s43-20260829T003332Z-3858bb98/trainer/checkpoint-1368",
|
| 5 |
+
"epoch": 8.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1368,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.14684287812041116,
|
| 14 |
+
"grad_norm": 8.596052169799805,
|
| 15 |
+
"learning_rate": 1.4035087719298246e-06,
|
| 16 |
+
"loss": 4.435,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.2936857562408223,
|
| 21 |
+
"grad_norm": 14.833935737609863,
|
| 22 |
+
"learning_rate": 2.865497076023392e-06,
|
| 23 |
+
"loss": 4.4063,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.44052863436123346,
|
| 28 |
+
"grad_norm": 13.407795906066895,
|
| 29 |
+
"learning_rate": 4.3274853801169596e-06,
|
| 30 |
+
"loss": 4.3999,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5873715124816447,
|
| 35 |
+
"grad_norm": 8.366018295288086,
|
| 36 |
+
"learning_rate": 5.789473684210527e-06,
|
| 37 |
+
"loss": 4.3939,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.7342143906020558,
|
| 42 |
+
"grad_norm": 10.895014762878418,
|
| 43 |
+
"learning_rate": 7.251461988304094e-06,
|
| 44 |
+
"loss": 4.3654,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.8810572687224669,
|
| 49 |
+
"grad_norm": 18.356342315673828,
|
| 50 |
+
"learning_rate": 8.713450292397661e-06,
|
| 51 |
+
"loss": 4.2785,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 1.0,
|
| 56 |
+
"eval_accuracy": 0.3790035587188612,
|
| 57 |
+
"eval_loss": 1.0731970071792603,
|
| 58 |
+
"eval_macro_f1": 0.27516202821197105,
|
| 59 |
+
"eval_runtime": 1.2374,
|
| 60 |
+
"eval_samples_per_second": 454.172,
|
| 61 |
+
"eval_steps_per_second": 29.093,
|
| 62 |
+
"eval_weighted_f1": 0.27619162134037395,
|
| 63 |
+
"step": 171
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 1.0234948604992657,
|
| 67 |
+
"grad_norm": 17.413288116455078,
|
| 68 |
+
"learning_rate": 1.017543859649123e-05,
|
| 69 |
+
"loss": 4.0857,
|
| 70 |
+
"step": 175
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.170337738619677,
|
| 74 |
+
"grad_norm": 27.80413818359375,
|
| 75 |
+
"learning_rate": 1.1637426900584797e-05,
|
| 76 |
+
"loss": 3.9363,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.3171806167400881,
|
| 81 |
+
"grad_norm": 50.66158676147461,
|
| 82 |
+
"learning_rate": 1.3099415204678362e-05,
|
| 83 |
+
"loss": 3.5504,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.4640234948604993,
|
| 88 |
+
"grad_norm": 46.63914489746094,
|
| 89 |
+
"learning_rate": 1.4561403508771931e-05,
|
| 90 |
+
"loss": 3.4581,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.6108663729809103,
|
| 95 |
+
"grad_norm": 60.495391845703125,
|
| 96 |
+
"learning_rate": 1.60233918128655e-05,
|
| 97 |
+
"loss": 3.0013,
|
| 98 |
+
"step": 275
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.7577092511013217,
|
| 102 |
+
"grad_norm": 62.41952896118164,
|
| 103 |
+
"learning_rate": 1.7485380116959064e-05,
|
| 104 |
+
"loss": 2.8489,
|
| 105 |
+
"step": 300
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.9045521292217327,
|
| 109 |
+
"grad_norm": 31.185758590698242,
|
| 110 |
+
"learning_rate": 1.894736842105263e-05,
|
| 111 |
+
"loss": 2.4566,
|
| 112 |
+
"step": 325
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 2.0,
|
| 116 |
+
"eval_accuracy": 0.7758007117437722,
|
| 117 |
+
"eval_loss": 0.5903128981590271,
|
| 118 |
+
"eval_macro_f1": 0.7616194301652798,
|
| 119 |
+
"eval_runtime": 1.2472,
|
| 120 |
+
"eval_samples_per_second": 450.605,
|
| 121 |
+
"eval_steps_per_second": 28.864,
|
| 122 |
+
"eval_weighted_f1": 0.7624637971800687,
|
| 123 |
+
"step": 342
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 2.0469897209985315,
|
| 127 |
+
"grad_norm": 54.89079284667969,
|
| 128 |
+
"learning_rate": 1.99545159194282e-05,
|
| 129 |
+
"loss": 2.1812,
|
| 130 |
+
"step": 350
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.193832599118943,
|
| 134 |
+
"grad_norm": 34.924903869628906,
|
| 135 |
+
"learning_rate": 1.9792072774528916e-05,
|
| 136 |
+
"loss": 2.0218,
|
| 137 |
+
"step": 375
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 2.340675477239354,
|
| 141 |
+
"grad_norm": 59.93123245239258,
|
| 142 |
+
"learning_rate": 1.962962962962963e-05,
|
| 143 |
+
"loss": 1.8863,
|
| 144 |
+
"step": 400
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.4875183553597653,
|
| 148 |
+
"grad_norm": 32.211692810058594,
|
| 149 |
+
"learning_rate": 1.9467186484730346e-05,
|
| 150 |
+
"loss": 1.9668,
|
| 151 |
+
"step": 425
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.6343612334801763,
|
| 155 |
+
"grad_norm": 18.865108489990234,
|
| 156 |
+
"learning_rate": 1.930474333983106e-05,
|
| 157 |
+
"loss": 2.0939,
|
| 158 |
+
"step": 450
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.7812041116005872,
|
| 162 |
+
"grad_norm": 52.83786392211914,
|
| 163 |
+
"learning_rate": 1.9142300194931776e-05,
|
| 164 |
+
"loss": 2.104,
|
| 165 |
+
"step": 475
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.9280469897209986,
|
| 169 |
+
"grad_norm": 40.296146392822266,
|
| 170 |
+
"learning_rate": 1.897985705003249e-05,
|
| 171 |
+
"loss": 1.9715,
|
| 172 |
+
"step": 500
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 3.0,
|
| 176 |
+
"eval_accuracy": 0.802491103202847,
|
| 177 |
+
"eval_loss": 0.5009528994560242,
|
| 178 |
+
"eval_macro_f1": 0.8050035970672421,
|
| 179 |
+
"eval_runtime": 1.0187,
|
| 180 |
+
"eval_samples_per_second": 551.678,
|
| 181 |
+
"eval_steps_per_second": 35.339,
|
| 182 |
+
"eval_weighted_f1": 0.8047108918825683,
|
| 183 |
+
"step": 513
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 3.0704845814977975,
|
| 187 |
+
"grad_norm": 53.98931884765625,
|
| 188 |
+
"learning_rate": 1.8817413905133203e-05,
|
| 189 |
+
"loss": 1.7326,
|
| 190 |
+
"step": 525
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 3.2173274596182084,
|
| 194 |
+
"grad_norm": 120.435791015625,
|
| 195 |
+
"learning_rate": 1.8654970760233918e-05,
|
| 196 |
+
"loss": 1.3275,
|
| 197 |
+
"step": 550
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 3.36417033773862,
|
| 201 |
+
"grad_norm": 44.686134338378906,
|
| 202 |
+
"learning_rate": 1.8492527615334633e-05,
|
| 203 |
+
"loss": 1.4738,
|
| 204 |
+
"step": 575
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.511013215859031,
|
| 208 |
+
"grad_norm": 34.26319122314453,
|
| 209 |
+
"learning_rate": 1.833008447043535e-05,
|
| 210 |
+
"loss": 1.3203,
|
| 211 |
+
"step": 600
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.657856093979442,
|
| 215 |
+
"grad_norm": 44.2707633972168,
|
| 216 |
+
"learning_rate": 1.8167641325536064e-05,
|
| 217 |
+
"loss": 1.2971,
|
| 218 |
+
"step": 625
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.804698972099853,
|
| 222 |
+
"grad_norm": 65.55018615722656,
|
| 223 |
+
"learning_rate": 1.800519818063678e-05,
|
| 224 |
+
"loss": 1.2726,
|
| 225 |
+
"step": 650
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.951541850220264,
|
| 229 |
+
"grad_norm": 44.38268280029297,
|
| 230 |
+
"learning_rate": 1.7842755035737494e-05,
|
| 231 |
+
"loss": 1.4155,
|
| 232 |
+
"step": 675
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 4.0,
|
| 236 |
+
"eval_accuracy": 0.806049822064057,
|
| 237 |
+
"eval_loss": 0.5667863488197327,
|
| 238 |
+
"eval_macro_f1": 0.8000615308437008,
|
| 239 |
+
"eval_runtime": 1.0115,
|
| 240 |
+
"eval_samples_per_second": 555.585,
|
| 241 |
+
"eval_steps_per_second": 35.589,
|
| 242 |
+
"eval_weighted_f1": 0.800242426134126,
|
| 243 |
+
"step": 684
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 4.093979441997063,
|
| 247 |
+
"grad_norm": 39.32135009765625,
|
| 248 |
+
"learning_rate": 1.768031189083821e-05,
|
| 249 |
+
"loss": 1.018,
|
| 250 |
+
"step": 700
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 4.240822320117474,
|
| 254 |
+
"grad_norm": 65.7984619140625,
|
| 255 |
+
"learning_rate": 1.751786874593892e-05,
|
| 256 |
+
"loss": 0.9702,
|
| 257 |
+
"step": 725
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 4.387665198237886,
|
| 261 |
+
"grad_norm": 46.16535568237305,
|
| 262 |
+
"learning_rate": 1.7355425601039636e-05,
|
| 263 |
+
"loss": 0.9768,
|
| 264 |
+
"step": 750
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.534508076358296,
|
| 268 |
+
"grad_norm": 44.18907928466797,
|
| 269 |
+
"learning_rate": 1.719298245614035e-05,
|
| 270 |
+
"loss": 1.0158,
|
| 271 |
+
"step": 775
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 4.681350954478708,
|
| 275 |
+
"grad_norm": 80.86709594726562,
|
| 276 |
+
"learning_rate": 1.7030539311241066e-05,
|
| 277 |
+
"loss": 1.032,
|
| 278 |
+
"step": 800
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 4.828193832599119,
|
| 282 |
+
"grad_norm": 48.729007720947266,
|
| 283 |
+
"learning_rate": 1.686809616634178e-05,
|
| 284 |
+
"loss": 0.9201,
|
| 285 |
+
"step": 825
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 4.975036710719531,
|
| 289 |
+
"grad_norm": 94.73574829101562,
|
| 290 |
+
"learning_rate": 1.6705653021442497e-05,
|
| 291 |
+
"loss": 0.9139,
|
| 292 |
+
"step": 850
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 5.0,
|
| 296 |
+
"eval_accuracy": 0.8185053380782918,
|
| 297 |
+
"eval_loss": 0.5922960042953491,
|
| 298 |
+
"eval_macro_f1": 0.8182962758895126,
|
| 299 |
+
"eval_runtime": 1.1924,
|
| 300 |
+
"eval_samples_per_second": 471.315,
|
| 301 |
+
"eval_steps_per_second": 30.191,
|
| 302 |
+
"eval_weighted_f1": 0.8182576376915619,
|
| 303 |
+
"step": 855
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 5.117474302496329,
|
| 307 |
+
"grad_norm": 49.79019546508789,
|
| 308 |
+
"learning_rate": 1.654320987654321e-05,
|
| 309 |
+
"loss": 0.7546,
|
| 310 |
+
"step": 875
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 5.26431718061674,
|
| 314 |
+
"grad_norm": 48.020660400390625,
|
| 315 |
+
"learning_rate": 1.6380766731643927e-05,
|
| 316 |
+
"loss": 0.524,
|
| 317 |
+
"step": 900
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 5.411160058737151,
|
| 321 |
+
"grad_norm": 38.63826370239258,
|
| 322 |
+
"learning_rate": 1.621832358674464e-05,
|
| 323 |
+
"loss": 0.6936,
|
| 324 |
+
"step": 925
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 5.558002936857562,
|
| 328 |
+
"grad_norm": 33.63285446166992,
|
| 329 |
+
"learning_rate": 1.6055880441845354e-05,
|
| 330 |
+
"loss": 0.6554,
|
| 331 |
+
"step": 950
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.704845814977974,
|
| 335 |
+
"grad_norm": 78.11824798583984,
|
| 336 |
+
"learning_rate": 1.589343729694607e-05,
|
| 337 |
+
"loss": 0.7186,
|
| 338 |
+
"step": 975
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 5.851688693098385,
|
| 342 |
+
"grad_norm": 24.796911239624023,
|
| 343 |
+
"learning_rate": 1.5730994152046787e-05,
|
| 344 |
+
"loss": 0.785,
|
| 345 |
+
"step": 1000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.998531571218796,
|
| 349 |
+
"grad_norm": 68.1156005859375,
|
| 350 |
+
"learning_rate": 1.55685510071475e-05,
|
| 351 |
+
"loss": 0.6549,
|
| 352 |
+
"step": 1025
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 6.0,
|
| 356 |
+
"eval_accuracy": 0.8167259786476868,
|
| 357 |
+
"eval_loss": 0.7059098482131958,
|
| 358 |
+
"eval_macro_f1": 0.815631087602807,
|
| 359 |
+
"eval_runtime": 1.0441,
|
| 360 |
+
"eval_samples_per_second": 538.271,
|
| 361 |
+
"eval_steps_per_second": 34.48,
|
| 362 |
+
"eval_weighted_f1": 0.8155098434679822,
|
| 363 |
+
"step": 1026
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 6.140969162995595,
|
| 367 |
+
"grad_norm": 44.33232116699219,
|
| 368 |
+
"learning_rate": 1.5406107862248214e-05,
|
| 369 |
+
"loss": 0.3111,
|
| 370 |
+
"step": 1050
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 6.2878120411160054,
|
| 374 |
+
"grad_norm": 31.627370834350586,
|
| 375 |
+
"learning_rate": 1.524366471734893e-05,
|
| 376 |
+
"loss": 0.415,
|
| 377 |
+
"step": 1075
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 6.434654919236417,
|
| 381 |
+
"grad_norm": 86.87671661376953,
|
| 382 |
+
"learning_rate": 1.5081221572449645e-05,
|
| 383 |
+
"loss": 0.6189,
|
| 384 |
+
"step": 1100
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 6.581497797356828,
|
| 388 |
+
"grad_norm": 27.572864532470703,
|
| 389 |
+
"learning_rate": 1.4918778427550358e-05,
|
| 390 |
+
"loss": 0.5276,
|
| 391 |
+
"step": 1125
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 6.72834067547724,
|
| 395 |
+
"grad_norm": 119.9407958984375,
|
| 396 |
+
"learning_rate": 1.4756335282651073e-05,
|
| 397 |
+
"loss": 0.5633,
|
| 398 |
+
"step": 1150
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 6.87518355359765,
|
| 402 |
+
"grad_norm": 170.39004516601562,
|
| 403 |
+
"learning_rate": 1.4593892137751788e-05,
|
| 404 |
+
"loss": 0.4878,
|
| 405 |
+
"step": 1175
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 7.0,
|
| 409 |
+
"eval_accuracy": 0.7846975088967971,
|
| 410 |
+
"eval_loss": 0.7756593227386475,
|
| 411 |
+
"eval_macro_f1": 0.787246552944017,
|
| 412 |
+
"eval_runtime": 1.0963,
|
| 413 |
+
"eval_samples_per_second": 512.616,
|
| 414 |
+
"eval_steps_per_second": 32.837,
|
| 415 |
+
"eval_weighted_f1": 0.7870545789248027,
|
| 416 |
+
"step": 1197
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 7.0176211453744495,
|
| 420 |
+
"grad_norm": 28.966751098632812,
|
| 421 |
+
"learning_rate": 1.4431448992852503e-05,
|
| 422 |
+
"loss": 0.6721,
|
| 423 |
+
"step": 1200
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 7.164464023494861,
|
| 427 |
+
"grad_norm": 27.362152099609375,
|
| 428 |
+
"learning_rate": 1.4269005847953217e-05,
|
| 429 |
+
"loss": 0.2625,
|
| 430 |
+
"step": 1225
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 7.311306901615271,
|
| 434 |
+
"grad_norm": 55.169097900390625,
|
| 435 |
+
"learning_rate": 1.4106562703053932e-05,
|
| 436 |
+
"loss": 0.4306,
|
| 437 |
+
"step": 1250
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 7.458149779735683,
|
| 441 |
+
"grad_norm": 53.57747268676758,
|
| 442 |
+
"learning_rate": 1.3944119558154647e-05,
|
| 443 |
+
"loss": 0.3237,
|
| 444 |
+
"step": 1275
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 7.604992657856094,
|
| 448 |
+
"grad_norm": 28.719289779663086,
|
| 449 |
+
"learning_rate": 1.3781676413255362e-05,
|
| 450 |
+
"loss": 0.1876,
|
| 451 |
+
"step": 1300
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 7.751835535976505,
|
| 455 |
+
"grad_norm": 183.29908752441406,
|
| 456 |
+
"learning_rate": 1.3619233268356076e-05,
|
| 457 |
+
"loss": 0.4213,
|
| 458 |
+
"step": 1325
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 7.898678414096916,
|
| 462 |
+
"grad_norm": 44.32665252685547,
|
| 463 |
+
"learning_rate": 1.345679012345679e-05,
|
| 464 |
+
"loss": 0.4004,
|
| 465 |
+
"step": 1350
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 8.0,
|
| 469 |
+
"eval_accuracy": 0.8256227758007118,
|
| 470 |
+
"eval_loss": 0.8886235356330872,
|
| 471 |
+
"eval_macro_f1": 0.824686944370819,
|
| 472 |
+
"eval_runtime": 1.1191,
|
| 473 |
+
"eval_samples_per_second": 502.171,
|
| 474 |
+
"eval_steps_per_second": 32.168,
|
| 475 |
+
"eval_weighted_f1": 0.824188749717917,
|
| 476 |
+
"step": 1368
|
| 477 |
+
}
|
| 478 |
+
],
|
| 479 |
+
"logging_steps": 25,
|
| 480 |
+
"max_steps": 3420,
|
| 481 |
+
"num_input_tokens_seen": 0,
|
| 482 |
+
"num_train_epochs": 20,
|
| 483 |
+
"save_steps": 500,
|
| 484 |
+
"stateful_callbacks": {
|
| 485 |
+
"EarlyStoppingCallback": {
|
| 486 |
+
"args": {
|
| 487 |
+
"early_stopping_patience": 3,
|
| 488 |
+
"early_stopping_threshold": 0.0
|
| 489 |
+
},
|
| 490 |
+
"attributes": {
|
| 491 |
+
"early_stopping_patience_counter": 0
|
| 492 |
+
}
|
| 493 |
+
},
|
| 494 |
+
"TrainerControl": {
|
| 495 |
+
"args": {
|
| 496 |
+
"should_epoch_stop": false,
|
| 497 |
+
"should_evaluate": false,
|
| 498 |
+
"should_log": false,
|
| 499 |
+
"should_save": true,
|
| 500 |
+
"should_training_stop": false
|
| 501 |
+
},
|
| 502 |
+
"attributes": {}
|
| 503 |
+
}
|
| 504 |
+
},
|
| 505 |
+
"total_flos": 0.0,
|
| 506 |
+
"train_batch_size": 8,
|
| 507 |
+
"trial_name": null,
|
| 508 |
+
"trial_params": null
|
| 509 |
+
}
|
ood-vifactcheck-bamibert-source-source-01-s43-20260829T003332Z-3858bb98/trainer/checkpoint-1881/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-source-source-01-s43-20260829T003332Z-3858bb98/trainer/checkpoint-1881/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-vifactcheck-bamibert-source-source-01-s43-20260829T003332Z-3858bb98/trainer/checkpoint-1881/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-source-source-01-s43-20260829T003332Z-3858bb98/trainer/checkpoint-1881/trainer_state.json
ADDED
|
@@ -0,0 +1,689 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1368,
|
| 3 |
+
"best_metric": 0.824686944370819,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-source-source-01-s43-20260829T003332Z-3858bb98/trainer/checkpoint-1368",
|
| 5 |
+
"epoch": 11.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1881,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.14684287812041116,
|
| 14 |
+
"grad_norm": 8.596052169799805,
|
| 15 |
+
"learning_rate": 1.4035087719298246e-06,
|
| 16 |
+
"loss": 4.435,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.2936857562408223,
|
| 21 |
+
"grad_norm": 14.833935737609863,
|
| 22 |
+
"learning_rate": 2.865497076023392e-06,
|
| 23 |
+
"loss": 4.4063,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.44052863436123346,
|
| 28 |
+
"grad_norm": 13.407795906066895,
|
| 29 |
+
"learning_rate": 4.3274853801169596e-06,
|
| 30 |
+
"loss": 4.3999,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5873715124816447,
|
| 35 |
+
"grad_norm": 8.366018295288086,
|
| 36 |
+
"learning_rate": 5.789473684210527e-06,
|
| 37 |
+
"loss": 4.3939,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.7342143906020558,
|
| 42 |
+
"grad_norm": 10.895014762878418,
|
| 43 |
+
"learning_rate": 7.251461988304094e-06,
|
| 44 |
+
"loss": 4.3654,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.8810572687224669,
|
| 49 |
+
"grad_norm": 18.356342315673828,
|
| 50 |
+
"learning_rate": 8.713450292397661e-06,
|
| 51 |
+
"loss": 4.2785,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 1.0,
|
| 56 |
+
"eval_accuracy": 0.3790035587188612,
|
| 57 |
+
"eval_loss": 1.0731970071792603,
|
| 58 |
+
"eval_macro_f1": 0.27516202821197105,
|
| 59 |
+
"eval_runtime": 1.2374,
|
| 60 |
+
"eval_samples_per_second": 454.172,
|
| 61 |
+
"eval_steps_per_second": 29.093,
|
| 62 |
+
"eval_weighted_f1": 0.27619162134037395,
|
| 63 |
+
"step": 171
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 1.0234948604992657,
|
| 67 |
+
"grad_norm": 17.413288116455078,
|
| 68 |
+
"learning_rate": 1.017543859649123e-05,
|
| 69 |
+
"loss": 4.0857,
|
| 70 |
+
"step": 175
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.170337738619677,
|
| 74 |
+
"grad_norm": 27.80413818359375,
|
| 75 |
+
"learning_rate": 1.1637426900584797e-05,
|
| 76 |
+
"loss": 3.9363,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.3171806167400881,
|
| 81 |
+
"grad_norm": 50.66158676147461,
|
| 82 |
+
"learning_rate": 1.3099415204678362e-05,
|
| 83 |
+
"loss": 3.5504,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.4640234948604993,
|
| 88 |
+
"grad_norm": 46.63914489746094,
|
| 89 |
+
"learning_rate": 1.4561403508771931e-05,
|
| 90 |
+
"loss": 3.4581,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.6108663729809103,
|
| 95 |
+
"grad_norm": 60.495391845703125,
|
| 96 |
+
"learning_rate": 1.60233918128655e-05,
|
| 97 |
+
"loss": 3.0013,
|
| 98 |
+
"step": 275
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.7577092511013217,
|
| 102 |
+
"grad_norm": 62.41952896118164,
|
| 103 |
+
"learning_rate": 1.7485380116959064e-05,
|
| 104 |
+
"loss": 2.8489,
|
| 105 |
+
"step": 300
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.9045521292217327,
|
| 109 |
+
"grad_norm": 31.185758590698242,
|
| 110 |
+
"learning_rate": 1.894736842105263e-05,
|
| 111 |
+
"loss": 2.4566,
|
| 112 |
+
"step": 325
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 2.0,
|
| 116 |
+
"eval_accuracy": 0.7758007117437722,
|
| 117 |
+
"eval_loss": 0.5903128981590271,
|
| 118 |
+
"eval_macro_f1": 0.7616194301652798,
|
| 119 |
+
"eval_runtime": 1.2472,
|
| 120 |
+
"eval_samples_per_second": 450.605,
|
| 121 |
+
"eval_steps_per_second": 28.864,
|
| 122 |
+
"eval_weighted_f1": 0.7624637971800687,
|
| 123 |
+
"step": 342
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 2.0469897209985315,
|
| 127 |
+
"grad_norm": 54.89079284667969,
|
| 128 |
+
"learning_rate": 1.99545159194282e-05,
|
| 129 |
+
"loss": 2.1812,
|
| 130 |
+
"step": 350
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.193832599118943,
|
| 134 |
+
"grad_norm": 34.924903869628906,
|
| 135 |
+
"learning_rate": 1.9792072774528916e-05,
|
| 136 |
+
"loss": 2.0218,
|
| 137 |
+
"step": 375
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 2.340675477239354,
|
| 141 |
+
"grad_norm": 59.93123245239258,
|
| 142 |
+
"learning_rate": 1.962962962962963e-05,
|
| 143 |
+
"loss": 1.8863,
|
| 144 |
+
"step": 400
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.4875183553597653,
|
| 148 |
+
"grad_norm": 32.211692810058594,
|
| 149 |
+
"learning_rate": 1.9467186484730346e-05,
|
| 150 |
+
"loss": 1.9668,
|
| 151 |
+
"step": 425
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.6343612334801763,
|
| 155 |
+
"grad_norm": 18.865108489990234,
|
| 156 |
+
"learning_rate": 1.930474333983106e-05,
|
| 157 |
+
"loss": 2.0939,
|
| 158 |
+
"step": 450
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.7812041116005872,
|
| 162 |
+
"grad_norm": 52.83786392211914,
|
| 163 |
+
"learning_rate": 1.9142300194931776e-05,
|
| 164 |
+
"loss": 2.104,
|
| 165 |
+
"step": 475
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.9280469897209986,
|
| 169 |
+
"grad_norm": 40.296146392822266,
|
| 170 |
+
"learning_rate": 1.897985705003249e-05,
|
| 171 |
+
"loss": 1.9715,
|
| 172 |
+
"step": 500
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 3.0,
|
| 176 |
+
"eval_accuracy": 0.802491103202847,
|
| 177 |
+
"eval_loss": 0.5009528994560242,
|
| 178 |
+
"eval_macro_f1": 0.8050035970672421,
|
| 179 |
+
"eval_runtime": 1.0187,
|
| 180 |
+
"eval_samples_per_second": 551.678,
|
| 181 |
+
"eval_steps_per_second": 35.339,
|
| 182 |
+
"eval_weighted_f1": 0.8047108918825683,
|
| 183 |
+
"step": 513
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 3.0704845814977975,
|
| 187 |
+
"grad_norm": 53.98931884765625,
|
| 188 |
+
"learning_rate": 1.8817413905133203e-05,
|
| 189 |
+
"loss": 1.7326,
|
| 190 |
+
"step": 525
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 3.2173274596182084,
|
| 194 |
+
"grad_norm": 120.435791015625,
|
| 195 |
+
"learning_rate": 1.8654970760233918e-05,
|
| 196 |
+
"loss": 1.3275,
|
| 197 |
+
"step": 550
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 3.36417033773862,
|
| 201 |
+
"grad_norm": 44.686134338378906,
|
| 202 |
+
"learning_rate": 1.8492527615334633e-05,
|
| 203 |
+
"loss": 1.4738,
|
| 204 |
+
"step": 575
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.511013215859031,
|
| 208 |
+
"grad_norm": 34.26319122314453,
|
| 209 |
+
"learning_rate": 1.833008447043535e-05,
|
| 210 |
+
"loss": 1.3203,
|
| 211 |
+
"step": 600
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.657856093979442,
|
| 215 |
+
"grad_norm": 44.2707633972168,
|
| 216 |
+
"learning_rate": 1.8167641325536064e-05,
|
| 217 |
+
"loss": 1.2971,
|
| 218 |
+
"step": 625
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.804698972099853,
|
| 222 |
+
"grad_norm": 65.55018615722656,
|
| 223 |
+
"learning_rate": 1.800519818063678e-05,
|
| 224 |
+
"loss": 1.2726,
|
| 225 |
+
"step": 650
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.951541850220264,
|
| 229 |
+
"grad_norm": 44.38268280029297,
|
| 230 |
+
"learning_rate": 1.7842755035737494e-05,
|
| 231 |
+
"loss": 1.4155,
|
| 232 |
+
"step": 675
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 4.0,
|
| 236 |
+
"eval_accuracy": 0.806049822064057,
|
| 237 |
+
"eval_loss": 0.5667863488197327,
|
| 238 |
+
"eval_macro_f1": 0.8000615308437008,
|
| 239 |
+
"eval_runtime": 1.0115,
|
| 240 |
+
"eval_samples_per_second": 555.585,
|
| 241 |
+
"eval_steps_per_second": 35.589,
|
| 242 |
+
"eval_weighted_f1": 0.800242426134126,
|
| 243 |
+
"step": 684
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 4.093979441997063,
|
| 247 |
+
"grad_norm": 39.32135009765625,
|
| 248 |
+
"learning_rate": 1.768031189083821e-05,
|
| 249 |
+
"loss": 1.018,
|
| 250 |
+
"step": 700
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 4.240822320117474,
|
| 254 |
+
"grad_norm": 65.7984619140625,
|
| 255 |
+
"learning_rate": 1.751786874593892e-05,
|
| 256 |
+
"loss": 0.9702,
|
| 257 |
+
"step": 725
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 4.387665198237886,
|
| 261 |
+
"grad_norm": 46.16535568237305,
|
| 262 |
+
"learning_rate": 1.7355425601039636e-05,
|
| 263 |
+
"loss": 0.9768,
|
| 264 |
+
"step": 750
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.534508076358296,
|
| 268 |
+
"grad_norm": 44.18907928466797,
|
| 269 |
+
"learning_rate": 1.719298245614035e-05,
|
| 270 |
+
"loss": 1.0158,
|
| 271 |
+
"step": 775
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 4.681350954478708,
|
| 275 |
+
"grad_norm": 80.86709594726562,
|
| 276 |
+
"learning_rate": 1.7030539311241066e-05,
|
| 277 |
+
"loss": 1.032,
|
| 278 |
+
"step": 800
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 4.828193832599119,
|
| 282 |
+
"grad_norm": 48.729007720947266,
|
| 283 |
+
"learning_rate": 1.686809616634178e-05,
|
| 284 |
+
"loss": 0.9201,
|
| 285 |
+
"step": 825
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 4.975036710719531,
|
| 289 |
+
"grad_norm": 94.73574829101562,
|
| 290 |
+
"learning_rate": 1.6705653021442497e-05,
|
| 291 |
+
"loss": 0.9139,
|
| 292 |
+
"step": 850
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 5.0,
|
| 296 |
+
"eval_accuracy": 0.8185053380782918,
|
| 297 |
+
"eval_loss": 0.5922960042953491,
|
| 298 |
+
"eval_macro_f1": 0.8182962758895126,
|
| 299 |
+
"eval_runtime": 1.1924,
|
| 300 |
+
"eval_samples_per_second": 471.315,
|
| 301 |
+
"eval_steps_per_second": 30.191,
|
| 302 |
+
"eval_weighted_f1": 0.8182576376915619,
|
| 303 |
+
"step": 855
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 5.117474302496329,
|
| 307 |
+
"grad_norm": 49.79019546508789,
|
| 308 |
+
"learning_rate": 1.654320987654321e-05,
|
| 309 |
+
"loss": 0.7546,
|
| 310 |
+
"step": 875
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 5.26431718061674,
|
| 314 |
+
"grad_norm": 48.020660400390625,
|
| 315 |
+
"learning_rate": 1.6380766731643927e-05,
|
| 316 |
+
"loss": 0.524,
|
| 317 |
+
"step": 900
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 5.411160058737151,
|
| 321 |
+
"grad_norm": 38.63826370239258,
|
| 322 |
+
"learning_rate": 1.621832358674464e-05,
|
| 323 |
+
"loss": 0.6936,
|
| 324 |
+
"step": 925
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 5.558002936857562,
|
| 328 |
+
"grad_norm": 33.63285446166992,
|
| 329 |
+
"learning_rate": 1.6055880441845354e-05,
|
| 330 |
+
"loss": 0.6554,
|
| 331 |
+
"step": 950
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.704845814977974,
|
| 335 |
+
"grad_norm": 78.11824798583984,
|
| 336 |
+
"learning_rate": 1.589343729694607e-05,
|
| 337 |
+
"loss": 0.7186,
|
| 338 |
+
"step": 975
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 5.851688693098385,
|
| 342 |
+
"grad_norm": 24.796911239624023,
|
| 343 |
+
"learning_rate": 1.5730994152046787e-05,
|
| 344 |
+
"loss": 0.785,
|
| 345 |
+
"step": 1000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.998531571218796,
|
| 349 |
+
"grad_norm": 68.1156005859375,
|
| 350 |
+
"learning_rate": 1.55685510071475e-05,
|
| 351 |
+
"loss": 0.6549,
|
| 352 |
+
"step": 1025
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 6.0,
|
| 356 |
+
"eval_accuracy": 0.8167259786476868,
|
| 357 |
+
"eval_loss": 0.7059098482131958,
|
| 358 |
+
"eval_macro_f1": 0.815631087602807,
|
| 359 |
+
"eval_runtime": 1.0441,
|
| 360 |
+
"eval_samples_per_second": 538.271,
|
| 361 |
+
"eval_steps_per_second": 34.48,
|
| 362 |
+
"eval_weighted_f1": 0.8155098434679822,
|
| 363 |
+
"step": 1026
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 6.140969162995595,
|
| 367 |
+
"grad_norm": 44.33232116699219,
|
| 368 |
+
"learning_rate": 1.5406107862248214e-05,
|
| 369 |
+
"loss": 0.3111,
|
| 370 |
+
"step": 1050
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 6.2878120411160054,
|
| 374 |
+
"grad_norm": 31.627370834350586,
|
| 375 |
+
"learning_rate": 1.524366471734893e-05,
|
| 376 |
+
"loss": 0.415,
|
| 377 |
+
"step": 1075
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 6.434654919236417,
|
| 381 |
+
"grad_norm": 86.87671661376953,
|
| 382 |
+
"learning_rate": 1.5081221572449645e-05,
|
| 383 |
+
"loss": 0.6189,
|
| 384 |
+
"step": 1100
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 6.581497797356828,
|
| 388 |
+
"grad_norm": 27.572864532470703,
|
| 389 |
+
"learning_rate": 1.4918778427550358e-05,
|
| 390 |
+
"loss": 0.5276,
|
| 391 |
+
"step": 1125
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 6.72834067547724,
|
| 395 |
+
"grad_norm": 119.9407958984375,
|
| 396 |
+
"learning_rate": 1.4756335282651073e-05,
|
| 397 |
+
"loss": 0.5633,
|
| 398 |
+
"step": 1150
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 6.87518355359765,
|
| 402 |
+
"grad_norm": 170.39004516601562,
|
| 403 |
+
"learning_rate": 1.4593892137751788e-05,
|
| 404 |
+
"loss": 0.4878,
|
| 405 |
+
"step": 1175
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 7.0,
|
| 409 |
+
"eval_accuracy": 0.7846975088967971,
|
| 410 |
+
"eval_loss": 0.7756593227386475,
|
| 411 |
+
"eval_macro_f1": 0.787246552944017,
|
| 412 |
+
"eval_runtime": 1.0963,
|
| 413 |
+
"eval_samples_per_second": 512.616,
|
| 414 |
+
"eval_steps_per_second": 32.837,
|
| 415 |
+
"eval_weighted_f1": 0.7870545789248027,
|
| 416 |
+
"step": 1197
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 7.0176211453744495,
|
| 420 |
+
"grad_norm": 28.966751098632812,
|
| 421 |
+
"learning_rate": 1.4431448992852503e-05,
|
| 422 |
+
"loss": 0.6721,
|
| 423 |
+
"step": 1200
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 7.164464023494861,
|
| 427 |
+
"grad_norm": 27.362152099609375,
|
| 428 |
+
"learning_rate": 1.4269005847953217e-05,
|
| 429 |
+
"loss": 0.2625,
|
| 430 |
+
"step": 1225
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 7.311306901615271,
|
| 434 |
+
"grad_norm": 55.169097900390625,
|
| 435 |
+
"learning_rate": 1.4106562703053932e-05,
|
| 436 |
+
"loss": 0.4306,
|
| 437 |
+
"step": 1250
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 7.458149779735683,
|
| 441 |
+
"grad_norm": 53.57747268676758,
|
| 442 |
+
"learning_rate": 1.3944119558154647e-05,
|
| 443 |
+
"loss": 0.3237,
|
| 444 |
+
"step": 1275
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 7.604992657856094,
|
| 448 |
+
"grad_norm": 28.719289779663086,
|
| 449 |
+
"learning_rate": 1.3781676413255362e-05,
|
| 450 |
+
"loss": 0.1876,
|
| 451 |
+
"step": 1300
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 7.751835535976505,
|
| 455 |
+
"grad_norm": 183.29908752441406,
|
| 456 |
+
"learning_rate": 1.3619233268356076e-05,
|
| 457 |
+
"loss": 0.4213,
|
| 458 |
+
"step": 1325
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 7.898678414096916,
|
| 462 |
+
"grad_norm": 44.32665252685547,
|
| 463 |
+
"learning_rate": 1.345679012345679e-05,
|
| 464 |
+
"loss": 0.4004,
|
| 465 |
+
"step": 1350
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 8.0,
|
| 469 |
+
"eval_accuracy": 0.8256227758007118,
|
| 470 |
+
"eval_loss": 0.8886235356330872,
|
| 471 |
+
"eval_macro_f1": 0.824686944370819,
|
| 472 |
+
"eval_runtime": 1.1191,
|
| 473 |
+
"eval_samples_per_second": 502.171,
|
| 474 |
+
"eval_steps_per_second": 32.168,
|
| 475 |
+
"eval_weighted_f1": 0.824188749717917,
|
| 476 |
+
"step": 1368
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 8.041116005873715,
|
| 480 |
+
"grad_norm": 3.1535863876342773,
|
| 481 |
+
"learning_rate": 1.3294346978557506e-05,
|
| 482 |
+
"loss": 0.3042,
|
| 483 |
+
"step": 1375
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 8.187958883994126,
|
| 487 |
+
"grad_norm": 0.40325185656547546,
|
| 488 |
+
"learning_rate": 1.3131903833658221e-05,
|
| 489 |
+
"loss": 0.2945,
|
| 490 |
+
"step": 1400
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 8.334801762114537,
|
| 494 |
+
"grad_norm": 34.86583709716797,
|
| 495 |
+
"learning_rate": 1.2969460688758935e-05,
|
| 496 |
+
"loss": 0.2426,
|
| 497 |
+
"step": 1425
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 8.481644640234949,
|
| 501 |
+
"grad_norm": 0.8419399261474609,
|
| 502 |
+
"learning_rate": 1.280701754385965e-05,
|
| 503 |
+
"loss": 0.2492,
|
| 504 |
+
"step": 1450
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 8.62848751835536,
|
| 508 |
+
"grad_norm": 0.08319421112537384,
|
| 509 |
+
"learning_rate": 1.2644574398960365e-05,
|
| 510 |
+
"loss": 0.0589,
|
| 511 |
+
"step": 1475
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 8.775330396475772,
|
| 515 |
+
"grad_norm": 0.3601745367050171,
|
| 516 |
+
"learning_rate": 1.248213125406108e-05,
|
| 517 |
+
"loss": 0.4341,
|
| 518 |
+
"step": 1500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 8.922173274596181,
|
| 522 |
+
"grad_norm": 114.1891098022461,
|
| 523 |
+
"learning_rate": 1.2319688109161793e-05,
|
| 524 |
+
"loss": 0.3276,
|
| 525 |
+
"step": 1525
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 9.0,
|
| 529 |
+
"eval_accuracy": 0.8167259786476868,
|
| 530 |
+
"eval_loss": 1.0970559120178223,
|
| 531 |
+
"eval_macro_f1": 0.8168261863424328,
|
| 532 |
+
"eval_runtime": 1.0454,
|
| 533 |
+
"eval_samples_per_second": 537.607,
|
| 534 |
+
"eval_steps_per_second": 34.437,
|
| 535 |
+
"eval_weighted_f1": 0.8165540909017893,
|
| 536 |
+
"step": 1539
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 9.06461086637298,
|
| 540 |
+
"grad_norm": 37.51606369018555,
|
| 541 |
+
"learning_rate": 1.2157244964262509e-05,
|
| 542 |
+
"loss": 0.3472,
|
| 543 |
+
"step": 1550
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 9.211453744493392,
|
| 547 |
+
"grad_norm": 3.579357385635376,
|
| 548 |
+
"learning_rate": 1.1994801819363224e-05,
|
| 549 |
+
"loss": 0.2318,
|
| 550 |
+
"step": 1575
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 9.358296622613803,
|
| 554 |
+
"grad_norm": 120.2245864868164,
|
| 555 |
+
"learning_rate": 1.1832358674463939e-05,
|
| 556 |
+
"loss": 0.1819,
|
| 557 |
+
"step": 1600
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 9.505139500734215,
|
| 561 |
+
"grad_norm": 32.90099334716797,
|
| 562 |
+
"learning_rate": 1.1669915529564652e-05,
|
| 563 |
+
"loss": 0.2068,
|
| 564 |
+
"step": 1625
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 9.651982378854626,
|
| 568 |
+
"grad_norm": 0.08505241572856903,
|
| 569 |
+
"learning_rate": 1.1507472384665367e-05,
|
| 570 |
+
"loss": 0.1863,
|
| 571 |
+
"step": 1650
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 9.798825256975036,
|
| 575 |
+
"grad_norm": 136.59449768066406,
|
| 576 |
+
"learning_rate": 1.1345029239766083e-05,
|
| 577 |
+
"loss": 0.1897,
|
| 578 |
+
"step": 1675
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 9.945668135095447,
|
| 582 |
+
"grad_norm": 10.790090560913086,
|
| 583 |
+
"learning_rate": 1.11825860948668e-05,
|
| 584 |
+
"loss": 0.2566,
|
| 585 |
+
"step": 1700
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 10.0,
|
| 589 |
+
"eval_accuracy": 0.8042704626334519,
|
| 590 |
+
"eval_loss": 1.2563985586166382,
|
| 591 |
+
"eval_macro_f1": 0.8020821982492463,
|
| 592 |
+
"eval_runtime": 1.0019,
|
| 593 |
+
"eval_samples_per_second": 560.914,
|
| 594 |
+
"eval_steps_per_second": 35.93,
|
| 595 |
+
"eval_weighted_f1": 0.8015060279508923,
|
| 596 |
+
"step": 1710
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 10.088105726872246,
|
| 600 |
+
"grad_norm": 0.5655809640884399,
|
| 601 |
+
"learning_rate": 1.1020142949967511e-05,
|
| 602 |
+
"loss": 0.2117,
|
| 603 |
+
"step": 1725
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 10.234948604992658,
|
| 607 |
+
"grad_norm": 127.04888153076172,
|
| 608 |
+
"learning_rate": 1.0857699805068226e-05,
|
| 609 |
+
"loss": 0.1417,
|
| 610 |
+
"step": 1750
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 10.38179148311307,
|
| 614 |
+
"grad_norm": 3.348532199859619,
|
| 615 |
+
"learning_rate": 1.0695256660168943e-05,
|
| 616 |
+
"loss": 0.0908,
|
| 617 |
+
"step": 1775
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 10.52863436123348,
|
| 621 |
+
"grad_norm": 295.4440612792969,
|
| 622 |
+
"learning_rate": 1.0532813515269658e-05,
|
| 623 |
+
"loss": 0.1495,
|
| 624 |
+
"step": 1800
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 10.675477239353892,
|
| 628 |
+
"grad_norm": 0.3164202868938446,
|
| 629 |
+
"learning_rate": 1.037037037037037e-05,
|
| 630 |
+
"loss": 0.2513,
|
| 631 |
+
"step": 1825
|
| 632 |
+
},
|
| 633 |
+
{
|
| 634 |
+
"epoch": 10.822320117474302,
|
| 635 |
+
"grad_norm": 0.053126927465200424,
|
| 636 |
+
"learning_rate": 1.0207927225471085e-05,
|
| 637 |
+
"loss": 0.1346,
|
| 638 |
+
"step": 1850
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 10.969162995594713,
|
| 642 |
+
"grad_norm": 0.7897791266441345,
|
| 643 |
+
"learning_rate": 1.0045484080571802e-05,
|
| 644 |
+
"loss": 0.1291,
|
| 645 |
+
"step": 1875
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 11.0,
|
| 649 |
+
"eval_accuracy": 0.8149466192170819,
|
| 650 |
+
"eval_loss": 1.1864032745361328,
|
| 651 |
+
"eval_macro_f1": 0.8139425105294499,
|
| 652 |
+
"eval_runtime": 0.9872,
|
| 653 |
+
"eval_samples_per_second": 569.28,
|
| 654 |
+
"eval_steps_per_second": 36.466,
|
| 655 |
+
"eval_weighted_f1": 0.8136199969406905,
|
| 656 |
+
"step": 1881
|
| 657 |
+
}
|
| 658 |
+
],
|
| 659 |
+
"logging_steps": 25,
|
| 660 |
+
"max_steps": 3420,
|
| 661 |
+
"num_input_tokens_seen": 0,
|
| 662 |
+
"num_train_epochs": 20,
|
| 663 |
+
"save_steps": 500,
|
| 664 |
+
"stateful_callbacks": {
|
| 665 |
+
"EarlyStoppingCallback": {
|
| 666 |
+
"args": {
|
| 667 |
+
"early_stopping_patience": 3,
|
| 668 |
+
"early_stopping_threshold": 0.0
|
| 669 |
+
},
|
| 670 |
+
"attributes": {
|
| 671 |
+
"early_stopping_patience_counter": 3
|
| 672 |
+
}
|
| 673 |
+
},
|
| 674 |
+
"TrainerControl": {
|
| 675 |
+
"args": {
|
| 676 |
+
"should_epoch_stop": false,
|
| 677 |
+
"should_evaluate": false,
|
| 678 |
+
"should_log": false,
|
| 679 |
+
"should_save": true,
|
| 680 |
+
"should_training_stop": true
|
| 681 |
+
},
|
| 682 |
+
"attributes": {}
|
| 683 |
+
}
|
| 684 |
+
},
|
| 685 |
+
"total_flos": 0.0,
|
| 686 |
+
"train_batch_size": 8,
|
| 687 |
+
"trial_name": null,
|
| 688 |
+
"trial_params": null
|
| 689 |
+
}
|
ood-vifactcheck-bamibert-source-source-02-s43-20260829T004338Z-3858bb98/trainer/checkpoint-2715/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-source-source-02-s43-20260829T004338Z-3858bb98/trainer/checkpoint-2715/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-vifactcheck-bamibert-source-source-02-s43-20260829T004338Z-3858bb98/trainer/checkpoint-2715/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-source-source-02-s43-20260829T004338Z-3858bb98/trainer/checkpoint-2715/trainer_state.json
ADDED
|
@@ -0,0 +1,964 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 2715,
|
| 3 |
+
"best_metric": 0.8450306048345264,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-source-source-02-s43-20260829T004338Z-3858bb98/trainer/checkpoint-2715",
|
| 5 |
+
"epoch": 15.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2715,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.13812154696132597,
|
| 14 |
+
"grad_norm": 11.385866165161133,
|
| 15 |
+
"learning_rate": 1.3259668508287293e-06,
|
| 16 |
+
"loss": 4.4148,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.27624309392265195,
|
| 21 |
+
"grad_norm": 8.704425811767578,
|
| 22 |
+
"learning_rate": 2.707182320441989e-06,
|
| 23 |
+
"loss": 4.4207,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.4143646408839779,
|
| 28 |
+
"grad_norm": 14.337090492248535,
|
| 29 |
+
"learning_rate": 4.088397790055249e-06,
|
| 30 |
+
"loss": 4.4151,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5524861878453039,
|
| 35 |
+
"grad_norm": 162.95692443847656,
|
| 36 |
+
"learning_rate": 5.469613259668509e-06,
|
| 37 |
+
"loss": 4.3881,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.6906077348066298,
|
| 42 |
+
"grad_norm": 12.436147689819336,
|
| 43 |
+
"learning_rate": 6.850828729281769e-06,
|
| 44 |
+
"loss": 4.3728,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.8287292817679558,
|
| 49 |
+
"grad_norm": 19.559032440185547,
|
| 50 |
+
"learning_rate": 8.232044198895029e-06,
|
| 51 |
+
"loss": 4.3198,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.9668508287292817,
|
| 56 |
+
"grad_norm": 20.87177085876465,
|
| 57 |
+
"learning_rate": 9.613259668508288e-06,
|
| 58 |
+
"loss": 4.3045,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 1.0,
|
| 63 |
+
"eval_accuracy": 0.48625429553264604,
|
| 64 |
+
"eval_loss": 1.0244207382202148,
|
| 65 |
+
"eval_macro_f1": 0.4643918568040181,
|
| 66 |
+
"eval_runtime": 1.1672,
|
| 67 |
+
"eval_samples_per_second": 498.609,
|
| 68 |
+
"eval_steps_per_second": 31.699,
|
| 69 |
+
"eval_weighted_f1": 0.46821140597701794,
|
| 70 |
+
"step": 181
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.1049723756906078,
|
| 74 |
+
"grad_norm": 38.50503921508789,
|
| 75 |
+
"learning_rate": 1.0994475138121547e-05,
|
| 76 |
+
"loss": 4.0763,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.2430939226519337,
|
| 81 |
+
"grad_norm": 32.16830062866211,
|
| 82 |
+
"learning_rate": 1.2375690607734806e-05,
|
| 83 |
+
"loss": 3.6891,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.3812154696132597,
|
| 88 |
+
"grad_norm": 57.03675842285156,
|
| 89 |
+
"learning_rate": 1.3756906077348067e-05,
|
| 90 |
+
"loss": 3.3411,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.5193370165745856,
|
| 95 |
+
"grad_norm": 59.425662994384766,
|
| 96 |
+
"learning_rate": 1.5138121546961326e-05,
|
| 97 |
+
"loss": 3.0682,
|
| 98 |
+
"step": 275
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.6574585635359116,
|
| 102 |
+
"grad_norm": 38.729461669921875,
|
| 103 |
+
"learning_rate": 1.6519337016574586e-05,
|
| 104 |
+
"loss": 2.9975,
|
| 105 |
+
"step": 300
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.7955801104972375,
|
| 109 |
+
"grad_norm": 65.91519165039062,
|
| 110 |
+
"learning_rate": 1.7900552486187846e-05,
|
| 111 |
+
"loss": 2.9397,
|
| 112 |
+
"step": 325
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 1.9337016574585635,
|
| 116 |
+
"grad_norm": 57.606727600097656,
|
| 117 |
+
"learning_rate": 1.9281767955801107e-05,
|
| 118 |
+
"loss": 2.5799,
|
| 119 |
+
"step": 350
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 2.0,
|
| 123 |
+
"eval_accuracy": 0.7422680412371134,
|
| 124 |
+
"eval_loss": 0.6076148152351379,
|
| 125 |
+
"eval_macro_f1": 0.729084090886048,
|
| 126 |
+
"eval_runtime": 1.138,
|
| 127 |
+
"eval_samples_per_second": 511.433,
|
| 128 |
+
"eval_steps_per_second": 32.514,
|
| 129 |
+
"eval_weighted_f1": 0.7309228992577754,
|
| 130 |
+
"step": 362
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.0718232044198897,
|
| 134 |
+
"grad_norm": 65.74261474609375,
|
| 135 |
+
"learning_rate": 1.9926335174953962e-05,
|
| 136 |
+
"loss": 2.3743,
|
| 137 |
+
"step": 375
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 2.2099447513812156,
|
| 141 |
+
"grad_norm": 66.28155517578125,
|
| 142 |
+
"learning_rate": 1.9772866789441377e-05,
|
| 143 |
+
"loss": 2.2332,
|
| 144 |
+
"step": 400
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.3480662983425415,
|
| 148 |
+
"grad_norm": 60.59415817260742,
|
| 149 |
+
"learning_rate": 1.9619398403928794e-05,
|
| 150 |
+
"loss": 2.2188,
|
| 151 |
+
"step": 425
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.4861878453038675,
|
| 155 |
+
"grad_norm": 80.37751770019531,
|
| 156 |
+
"learning_rate": 1.946593001841621e-05,
|
| 157 |
+
"loss": 2.1052,
|
| 158 |
+
"step": 450
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.6243093922651934,
|
| 162 |
+
"grad_norm": 53.990535736083984,
|
| 163 |
+
"learning_rate": 1.9312461632903623e-05,
|
| 164 |
+
"loss": 2.0804,
|
| 165 |
+
"step": 475
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.7624309392265194,
|
| 169 |
+
"grad_norm": 38.29616928100586,
|
| 170 |
+
"learning_rate": 1.915899324739104e-05,
|
| 171 |
+
"loss": 1.9099,
|
| 172 |
+
"step": 500
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 2.9005524861878453,
|
| 176 |
+
"grad_norm": 26.27406883239746,
|
| 177 |
+
"learning_rate": 1.9005524861878455e-05,
|
| 178 |
+
"loss": 1.9504,
|
| 179 |
+
"step": 525
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 3.0,
|
| 183 |
+
"eval_accuracy": 0.7800687285223368,
|
| 184 |
+
"eval_loss": 0.5324428081512451,
|
| 185 |
+
"eval_macro_f1": 0.7815604168357787,
|
| 186 |
+
"eval_runtime": 1.1299,
|
| 187 |
+
"eval_samples_per_second": 515.079,
|
| 188 |
+
"eval_steps_per_second": 32.746,
|
| 189 |
+
"eval_weighted_f1": 0.7826158434225067,
|
| 190 |
+
"step": 543
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 3.0386740331491713,
|
| 194 |
+
"grad_norm": 36.635494232177734,
|
| 195 |
+
"learning_rate": 1.885205647636587e-05,
|
| 196 |
+
"loss": 1.9231,
|
| 197 |
+
"step": 550
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 3.1767955801104972,
|
| 201 |
+
"grad_norm": 38.11379623413086,
|
| 202 |
+
"learning_rate": 1.8698588090853287e-05,
|
| 203 |
+
"loss": 1.4116,
|
| 204 |
+
"step": 575
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.314917127071823,
|
| 208 |
+
"grad_norm": 31.0555477142334,
|
| 209 |
+
"learning_rate": 1.85451197053407e-05,
|
| 210 |
+
"loss": 1.7786,
|
| 211 |
+
"step": 600
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.453038674033149,
|
| 215 |
+
"grad_norm": 62.93512725830078,
|
| 216 |
+
"learning_rate": 1.8391651319828116e-05,
|
| 217 |
+
"loss": 1.462,
|
| 218 |
+
"step": 625
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.591160220994475,
|
| 222 |
+
"grad_norm": 44.05408477783203,
|
| 223 |
+
"learning_rate": 1.823818293431553e-05,
|
| 224 |
+
"loss": 1.5503,
|
| 225 |
+
"step": 650
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.729281767955801,
|
| 229 |
+
"grad_norm": 27.487030029296875,
|
| 230 |
+
"learning_rate": 1.8084714548802948e-05,
|
| 231 |
+
"loss": 1.399,
|
| 232 |
+
"step": 675
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.867403314917127,
|
| 236 |
+
"grad_norm": 41.88737487792969,
|
| 237 |
+
"learning_rate": 1.7931246163290362e-05,
|
| 238 |
+
"loss": 1.4536,
|
| 239 |
+
"step": 700
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 4.0,
|
| 243 |
+
"eval_accuracy": 0.7955326460481099,
|
| 244 |
+
"eval_loss": 0.591875433921814,
|
| 245 |
+
"eval_macro_f1": 0.795489967546203,
|
| 246 |
+
"eval_runtime": 1.0189,
|
| 247 |
+
"eval_samples_per_second": 571.178,
|
| 248 |
+
"eval_steps_per_second": 36.312,
|
| 249 |
+
"eval_weighted_f1": 0.7956920276233653,
|
| 250 |
+
"step": 724
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 4.005524861878453,
|
| 254 |
+
"grad_norm": 40.481544494628906,
|
| 255 |
+
"learning_rate": 1.7777777777777777e-05,
|
| 256 |
+
"loss": 1.5514,
|
| 257 |
+
"step": 725
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 4.143646408839779,
|
| 261 |
+
"grad_norm": 62.872493743896484,
|
| 262 |
+
"learning_rate": 1.7624309392265194e-05,
|
| 263 |
+
"loss": 0.8882,
|
| 264 |
+
"step": 750
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.281767955801105,
|
| 268 |
+
"grad_norm": 31.99635887145996,
|
| 269 |
+
"learning_rate": 1.747084100675261e-05,
|
| 270 |
+
"loss": 0.917,
|
| 271 |
+
"step": 775
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 4.419889502762431,
|
| 275 |
+
"grad_norm": 95.87877655029297,
|
| 276 |
+
"learning_rate": 1.7317372621240026e-05,
|
| 277 |
+
"loss": 1.0126,
|
| 278 |
+
"step": 800
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 4.558011049723757,
|
| 282 |
+
"grad_norm": 20.457937240600586,
|
| 283 |
+
"learning_rate": 1.716390423572744e-05,
|
| 284 |
+
"loss": 0.9183,
|
| 285 |
+
"step": 825
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 4.696132596685083,
|
| 289 |
+
"grad_norm": 69.46328735351562,
|
| 290 |
+
"learning_rate": 1.701043585021486e-05,
|
| 291 |
+
"loss": 0.8732,
|
| 292 |
+
"step": 850
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 4.834254143646409,
|
| 296 |
+
"grad_norm": 43.254112243652344,
|
| 297 |
+
"learning_rate": 1.6856967464702273e-05,
|
| 298 |
+
"loss": 0.9655,
|
| 299 |
+
"step": 875
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 4.972375690607735,
|
| 303 |
+
"grad_norm": 57.20595932006836,
|
| 304 |
+
"learning_rate": 1.670349907918969e-05,
|
| 305 |
+
"loss": 1.0731,
|
| 306 |
+
"step": 900
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 5.0,
|
| 310 |
+
"eval_accuracy": 0.802405498281787,
|
| 311 |
+
"eval_loss": 0.5767754316329956,
|
| 312 |
+
"eval_macro_f1": 0.8012312938988065,
|
| 313 |
+
"eval_runtime": 1.1082,
|
| 314 |
+
"eval_samples_per_second": 525.19,
|
| 315 |
+
"eval_steps_per_second": 33.388,
|
| 316 |
+
"eval_weighted_f1": 0.8020146931073064,
|
| 317 |
+
"step": 905
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 5.110497237569061,
|
| 321 |
+
"grad_norm": 30.346569061279297,
|
| 322 |
+
"learning_rate": 1.6550030693677105e-05,
|
| 323 |
+
"loss": 0.6757,
|
| 324 |
+
"step": 925
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 5.248618784530387,
|
| 328 |
+
"grad_norm": 79.27243041992188,
|
| 329 |
+
"learning_rate": 1.639656230816452e-05,
|
| 330 |
+
"loss": 0.7229,
|
| 331 |
+
"step": 950
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.386740331491713,
|
| 335 |
+
"grad_norm": 76.34603118896484,
|
| 336 |
+
"learning_rate": 1.6243093922651937e-05,
|
| 337 |
+
"loss": 0.7669,
|
| 338 |
+
"step": 975
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 5.524861878453039,
|
| 342 |
+
"grad_norm": 39.100955963134766,
|
| 343 |
+
"learning_rate": 1.608962553713935e-05,
|
| 344 |
+
"loss": 0.7318,
|
| 345 |
+
"step": 1000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.662983425414365,
|
| 349 |
+
"grad_norm": 44.392601013183594,
|
| 350 |
+
"learning_rate": 1.5936157151626766e-05,
|
| 351 |
+
"loss": 0.6025,
|
| 352 |
+
"step": 1025
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 5.801104972375691,
|
| 356 |
+
"grad_norm": 76.7225112915039,
|
| 357 |
+
"learning_rate": 1.5782688766114183e-05,
|
| 358 |
+
"loss": 0.6433,
|
| 359 |
+
"step": 1050
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 5.939226519337017,
|
| 363 |
+
"grad_norm": 26.3995361328125,
|
| 364 |
+
"learning_rate": 1.5629220380601598e-05,
|
| 365 |
+
"loss": 0.7356,
|
| 366 |
+
"step": 1075
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 6.0,
|
| 370 |
+
"eval_accuracy": 0.8041237113402062,
|
| 371 |
+
"eval_loss": 0.755528450012207,
|
| 372 |
+
"eval_macro_f1": 0.8035958989131342,
|
| 373 |
+
"eval_runtime": 1.1151,
|
| 374 |
+
"eval_samples_per_second": 521.931,
|
| 375 |
+
"eval_steps_per_second": 33.181,
|
| 376 |
+
"eval_weighted_f1": 0.8039718881665117,
|
| 377 |
+
"step": 1086
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 6.077348066298343,
|
| 381 |
+
"grad_norm": 16.3472900390625,
|
| 382 |
+
"learning_rate": 1.5475751995089012e-05,
|
| 383 |
+
"loss": 0.7126,
|
| 384 |
+
"step": 1100
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 6.2154696132596685,
|
| 388 |
+
"grad_norm": 67.30068969726562,
|
| 389 |
+
"learning_rate": 1.532228360957643e-05,
|
| 390 |
+
"loss": 0.4377,
|
| 391 |
+
"step": 1125
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 6.3535911602209945,
|
| 395 |
+
"grad_norm": 15.90624713897705,
|
| 396 |
+
"learning_rate": 1.5168815224063844e-05,
|
| 397 |
+
"loss": 0.4894,
|
| 398 |
+
"step": 1150
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 6.49171270718232,
|
| 402 |
+
"grad_norm": 119.05094146728516,
|
| 403 |
+
"learning_rate": 1.5015346838551258e-05,
|
| 404 |
+
"loss": 0.5446,
|
| 405 |
+
"step": 1175
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 6.629834254143646,
|
| 409 |
+
"grad_norm": 6.7117156982421875,
|
| 410 |
+
"learning_rate": 1.4861878453038676e-05,
|
| 411 |
+
"loss": 0.4783,
|
| 412 |
+
"step": 1200
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 6.767955801104972,
|
| 416 |
+
"grad_norm": 81.19427490234375,
|
| 417 |
+
"learning_rate": 1.470841006752609e-05,
|
| 418 |
+
"loss": 0.5304,
|
| 419 |
+
"step": 1225
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 6.906077348066298,
|
| 423 |
+
"grad_norm": 71.07677459716797,
|
| 424 |
+
"learning_rate": 1.4554941682013506e-05,
|
| 425 |
+
"loss": 0.4676,
|
| 426 |
+
"step": 1250
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 7.0,
|
| 430 |
+
"eval_accuracy": 0.8127147766323024,
|
| 431 |
+
"eval_loss": 0.8740180134773254,
|
| 432 |
+
"eval_macro_f1": 0.8119854921936865,
|
| 433 |
+
"eval_runtime": 1.0605,
|
| 434 |
+
"eval_samples_per_second": 548.792,
|
| 435 |
+
"eval_steps_per_second": 34.889,
|
| 436 |
+
"eval_weighted_f1": 0.8123917469310626,
|
| 437 |
+
"step": 1267
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 7.044198895027624,
|
| 441 |
+
"grad_norm": 3.1372318267822266,
|
| 442 |
+
"learning_rate": 1.4401473296500922e-05,
|
| 443 |
+
"loss": 0.4766,
|
| 444 |
+
"step": 1275
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 7.18232044198895,
|
| 448 |
+
"grad_norm": 24.158937454223633,
|
| 449 |
+
"learning_rate": 1.4248004910988338e-05,
|
| 450 |
+
"loss": 0.3425,
|
| 451 |
+
"step": 1300
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 7.320441988950276,
|
| 455 |
+
"grad_norm": 40.12568664550781,
|
| 456 |
+
"learning_rate": 1.4094536525475753e-05,
|
| 457 |
+
"loss": 0.3982,
|
| 458 |
+
"step": 1325
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 7.458563535911602,
|
| 462 |
+
"grad_norm": 105.64920043945312,
|
| 463 |
+
"learning_rate": 1.394106813996317e-05,
|
| 464 |
+
"loss": 0.4014,
|
| 465 |
+
"step": 1350
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 7.596685082872928,
|
| 469 |
+
"grad_norm": 17.37604522705078,
|
| 470 |
+
"learning_rate": 1.3787599754450585e-05,
|
| 471 |
+
"loss": 0.4638,
|
| 472 |
+
"step": 1375
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 7.734806629834254,
|
| 476 |
+
"grad_norm": 113.165283203125,
|
| 477 |
+
"learning_rate": 1.3634131368937999e-05,
|
| 478 |
+
"loss": 0.3836,
|
| 479 |
+
"step": 1400
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 7.87292817679558,
|
| 483 |
+
"grad_norm": 49.570980072021484,
|
| 484 |
+
"learning_rate": 1.3480662983425417e-05,
|
| 485 |
+
"loss": 0.434,
|
| 486 |
+
"step": 1425
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 8.0,
|
| 490 |
+
"eval_accuracy": 0.8075601374570447,
|
| 491 |
+
"eval_loss": 1.010019302368164,
|
| 492 |
+
"eval_macro_f1": 0.8053691385981757,
|
| 493 |
+
"eval_runtime": 1.0948,
|
| 494 |
+
"eval_samples_per_second": 531.611,
|
| 495 |
+
"eval_steps_per_second": 33.797,
|
| 496 |
+
"eval_weighted_f1": 0.8064679133325544,
|
| 497 |
+
"step": 1448
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 8.011049723756907,
|
| 501 |
+
"grad_norm": 41.95655059814453,
|
| 502 |
+
"learning_rate": 1.3327194597912831e-05,
|
| 503 |
+
"loss": 0.4402,
|
| 504 |
+
"step": 1450
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 8.149171270718233,
|
| 508 |
+
"grad_norm": 37.319374084472656,
|
| 509 |
+
"learning_rate": 1.3173726212400245e-05,
|
| 510 |
+
"loss": 0.192,
|
| 511 |
+
"step": 1475
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 8.287292817679559,
|
| 515 |
+
"grad_norm": 5.263917922973633,
|
| 516 |
+
"learning_rate": 1.3020257826887662e-05,
|
| 517 |
+
"loss": 0.3384,
|
| 518 |
+
"step": 1500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 8.425414364640885,
|
| 522 |
+
"grad_norm": 4.049027919769287,
|
| 523 |
+
"learning_rate": 1.2866789441375078e-05,
|
| 524 |
+
"loss": 0.3778,
|
| 525 |
+
"step": 1525
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 8.56353591160221,
|
| 529 |
+
"grad_norm": 134.29124450683594,
|
| 530 |
+
"learning_rate": 1.2713321055862494e-05,
|
| 531 |
+
"loss": 0.3379,
|
| 532 |
+
"step": 1550
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 8.701657458563536,
|
| 536 |
+
"grad_norm": 118.86800384521484,
|
| 537 |
+
"learning_rate": 1.2559852670349908e-05,
|
| 538 |
+
"loss": 0.2676,
|
| 539 |
+
"step": 1575
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 8.839779005524862,
|
| 543 |
+
"grad_norm": 0.21576865017414093,
|
| 544 |
+
"learning_rate": 1.2406384284837326e-05,
|
| 545 |
+
"loss": 0.1817,
|
| 546 |
+
"step": 1600
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 8.977900552486188,
|
| 550 |
+
"grad_norm": 266.0789794921875,
|
| 551 |
+
"learning_rate": 1.225291589932474e-05,
|
| 552 |
+
"loss": 0.437,
|
| 553 |
+
"step": 1625
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 9.0,
|
| 557 |
+
"eval_accuracy": 0.8230240549828178,
|
| 558 |
+
"eval_loss": 1.01772940158844,
|
| 559 |
+
"eval_macro_f1": 0.8219837928724036,
|
| 560 |
+
"eval_runtime": 1.0465,
|
| 561 |
+
"eval_samples_per_second": 556.148,
|
| 562 |
+
"eval_steps_per_second": 35.356,
|
| 563 |
+
"eval_weighted_f1": 0.8225489622623926,
|
| 564 |
+
"step": 1629
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 9.116022099447514,
|
| 568 |
+
"grad_norm": 6.092472076416016,
|
| 569 |
+
"learning_rate": 1.2099447513812154e-05,
|
| 570 |
+
"loss": 0.2903,
|
| 571 |
+
"step": 1650
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 9.25414364640884,
|
| 575 |
+
"grad_norm": 3.596750497817993,
|
| 576 |
+
"learning_rate": 1.1945979128299572e-05,
|
| 577 |
+
"loss": 0.2076,
|
| 578 |
+
"step": 1675
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 9.392265193370166,
|
| 582 |
+
"grad_norm": 2.6152288913726807,
|
| 583 |
+
"learning_rate": 1.1792510742786986e-05,
|
| 584 |
+
"loss": 0.3777,
|
| 585 |
+
"step": 1700
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 9.530386740331492,
|
| 589 |
+
"grad_norm": 5.998469352722168,
|
| 590 |
+
"learning_rate": 1.1639042357274402e-05,
|
| 591 |
+
"loss": 0.2567,
|
| 592 |
+
"step": 1725
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 9.668508287292818,
|
| 596 |
+
"grad_norm": 63.65530776977539,
|
| 597 |
+
"learning_rate": 1.1485573971761818e-05,
|
| 598 |
+
"loss": 0.3461,
|
| 599 |
+
"step": 1750
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 9.806629834254144,
|
| 603 |
+
"grad_norm": 289.3068542480469,
|
| 604 |
+
"learning_rate": 1.1332105586249233e-05,
|
| 605 |
+
"loss": 0.2654,
|
| 606 |
+
"step": 1775
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 9.94475138121547,
|
| 610 |
+
"grad_norm": 0.29069459438323975,
|
| 611 |
+
"learning_rate": 1.1178637200736649e-05,
|
| 612 |
+
"loss": 0.2365,
|
| 613 |
+
"step": 1800
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 10.0,
|
| 617 |
+
"eval_accuracy": 0.8230240549828178,
|
| 618 |
+
"eval_loss": 1.0735164880752563,
|
| 619 |
+
"eval_macro_f1": 0.8211941595494235,
|
| 620 |
+
"eval_runtime": 1.1903,
|
| 621 |
+
"eval_samples_per_second": 488.952,
|
| 622 |
+
"eval_steps_per_second": 31.085,
|
| 623 |
+
"eval_weighted_f1": 0.822216029780875,
|
| 624 |
+
"step": 1810
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 10.082872928176796,
|
| 628 |
+
"grad_norm": 0.36485153436660767,
|
| 629 |
+
"learning_rate": 1.1025168815224065e-05,
|
| 630 |
+
"loss": 0.1739,
|
| 631 |
+
"step": 1825
|
| 632 |
+
},
|
| 633 |
+
{
|
| 634 |
+
"epoch": 10.220994475138122,
|
| 635 |
+
"grad_norm": 0.6285772323608398,
|
| 636 |
+
"learning_rate": 1.087170042971148e-05,
|
| 637 |
+
"loss": 0.1638,
|
| 638 |
+
"step": 1850
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 10.359116022099448,
|
| 642 |
+
"grad_norm": 212.32395935058594,
|
| 643 |
+
"learning_rate": 1.0718232044198895e-05,
|
| 644 |
+
"loss": 0.2131,
|
| 645 |
+
"step": 1875
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 10.497237569060774,
|
| 649 |
+
"grad_norm": 265.7480773925781,
|
| 650 |
+
"learning_rate": 1.0564763658686313e-05,
|
| 651 |
+
"loss": 0.2143,
|
| 652 |
+
"step": 1900
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"epoch": 10.6353591160221,
|
| 656 |
+
"grad_norm": 0.04942381754517555,
|
| 657 |
+
"learning_rate": 1.0411295273173727e-05,
|
| 658 |
+
"loss": 0.1551,
|
| 659 |
+
"step": 1925
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 10.773480662983426,
|
| 663 |
+
"grad_norm": 39.93251037597656,
|
| 664 |
+
"learning_rate": 1.0257826887661141e-05,
|
| 665 |
+
"loss": 0.1928,
|
| 666 |
+
"step": 1950
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 10.911602209944752,
|
| 670 |
+
"grad_norm": 0.06080511212348938,
|
| 671 |
+
"learning_rate": 1.010435850214856e-05,
|
| 672 |
+
"loss": 0.1907,
|
| 673 |
+
"step": 1975
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 11.0,
|
| 677 |
+
"eval_accuracy": 0.8264604810996563,
|
| 678 |
+
"eval_loss": 1.064011573791504,
|
| 679 |
+
"eval_macro_f1": 0.8259139762802074,
|
| 680 |
+
"eval_runtime": 1.015,
|
| 681 |
+
"eval_samples_per_second": 573.375,
|
| 682 |
+
"eval_steps_per_second": 36.452,
|
| 683 |
+
"eval_weighted_f1": 0.8267390565929533,
|
| 684 |
+
"step": 1991
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 11.049723756906078,
|
| 688 |
+
"grad_norm": 1.0326106548309326,
|
| 689 |
+
"learning_rate": 9.950890116635974e-06,
|
| 690 |
+
"loss": 0.1865,
|
| 691 |
+
"step": 2000
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 11.187845303867404,
|
| 695 |
+
"grad_norm": 149.4682159423828,
|
| 696 |
+
"learning_rate": 9.79742173112339e-06,
|
| 697 |
+
"loss": 0.1998,
|
| 698 |
+
"step": 2025
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 11.32596685082873,
|
| 702 |
+
"grad_norm": 0.06841234862804413,
|
| 703 |
+
"learning_rate": 9.643953345610806e-06,
|
| 704 |
+
"loss": 0.1445,
|
| 705 |
+
"step": 2050
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 11.464088397790055,
|
| 709 |
+
"grad_norm": 25.998315811157227,
|
| 710 |
+
"learning_rate": 9.49048496009822e-06,
|
| 711 |
+
"loss": 0.1885,
|
| 712 |
+
"step": 2075
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 11.602209944751381,
|
| 716 |
+
"grad_norm": 0.6273815035820007,
|
| 717 |
+
"learning_rate": 9.337016574585636e-06,
|
| 718 |
+
"loss": 0.2335,
|
| 719 |
+
"step": 2100
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 11.740331491712707,
|
| 723 |
+
"grad_norm": 0.15333878993988037,
|
| 724 |
+
"learning_rate": 9.183548189073052e-06,
|
| 725 |
+
"loss": 0.1139,
|
| 726 |
+
"step": 2125
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 11.878453038674033,
|
| 730 |
+
"grad_norm": 0.04380929842591286,
|
| 731 |
+
"learning_rate": 9.030079803560468e-06,
|
| 732 |
+
"loss": 0.1818,
|
| 733 |
+
"step": 2150
|
| 734 |
+
},
|
| 735 |
+
{
|
| 736 |
+
"epoch": 12.0,
|
| 737 |
+
"eval_accuracy": 0.8281786941580757,
|
| 738 |
+
"eval_loss": 1.1222890615463257,
|
| 739 |
+
"eval_macro_f1": 0.8271145219966799,
|
| 740 |
+
"eval_runtime": 1.1021,
|
| 741 |
+
"eval_samples_per_second": 528.097,
|
| 742 |
+
"eval_steps_per_second": 33.573,
|
| 743 |
+
"eval_weighted_f1": 0.8276554250282702,
|
| 744 |
+
"step": 2172
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 12.01657458563536,
|
| 748 |
+
"grad_norm": 0.09283286333084106,
|
| 749 |
+
"learning_rate": 8.876611418047882e-06,
|
| 750 |
+
"loss": 0.2025,
|
| 751 |
+
"step": 2175
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 12.154696132596685,
|
| 755 |
+
"grad_norm": 0.03827335685491562,
|
| 756 |
+
"learning_rate": 8.723143032535298e-06,
|
| 757 |
+
"loss": 0.0699,
|
| 758 |
+
"step": 2200
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 12.292817679558011,
|
| 762 |
+
"grad_norm": 0.24005241692066193,
|
| 763 |
+
"learning_rate": 8.569674647022714e-06,
|
| 764 |
+
"loss": 0.1297,
|
| 765 |
+
"step": 2225
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 12.430939226519337,
|
| 769 |
+
"grad_norm": 3.1738510131835938,
|
| 770 |
+
"learning_rate": 8.416206261510129e-06,
|
| 771 |
+
"loss": 0.0815,
|
| 772 |
+
"step": 2250
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 12.569060773480663,
|
| 776 |
+
"grad_norm": 25.839738845825195,
|
| 777 |
+
"learning_rate": 8.262737875997545e-06,
|
| 778 |
+
"loss": 0.1777,
|
| 779 |
+
"step": 2275
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 12.707182320441989,
|
| 783 |
+
"grad_norm": 0.046633850783109665,
|
| 784 |
+
"learning_rate": 8.10926949048496e-06,
|
| 785 |
+
"loss": 0.0432,
|
| 786 |
+
"step": 2300
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 12.845303867403315,
|
| 790 |
+
"grad_norm": 486.20989990234375,
|
| 791 |
+
"learning_rate": 7.955801104972377e-06,
|
| 792 |
+
"loss": 0.0231,
|
| 793 |
+
"step": 2325
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 12.98342541436464,
|
| 797 |
+
"grad_norm": 0.04107821732759476,
|
| 798 |
+
"learning_rate": 7.802332719459793e-06,
|
| 799 |
+
"loss": 0.1274,
|
| 800 |
+
"step": 2350
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 13.0,
|
| 804 |
+
"eval_accuracy": 0.8298969072164949,
|
| 805 |
+
"eval_loss": 1.1414563655853271,
|
| 806 |
+
"eval_macro_f1": 0.8294064887271121,
|
| 807 |
+
"eval_runtime": 1.2772,
|
| 808 |
+
"eval_samples_per_second": 455.682,
|
| 809 |
+
"eval_steps_per_second": 28.969,
|
| 810 |
+
"eval_weighted_f1": 0.8298931070542709,
|
| 811 |
+
"step": 2353
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"epoch": 13.121546961325967,
|
| 815 |
+
"grad_norm": 354.08575439453125,
|
| 816 |
+
"learning_rate": 7.648864333947207e-06,
|
| 817 |
+
"loss": 0.1001,
|
| 818 |
+
"step": 2375
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 13.259668508287293,
|
| 822 |
+
"grad_norm": 0.16044709086418152,
|
| 823 |
+
"learning_rate": 7.495395948434623e-06,
|
| 824 |
+
"loss": 0.0414,
|
| 825 |
+
"step": 2400
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 13.397790055248619,
|
| 829 |
+
"grad_norm": 0.04437292739748955,
|
| 830 |
+
"learning_rate": 7.341927562922039e-06,
|
| 831 |
+
"loss": 0.0713,
|
| 832 |
+
"step": 2425
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 13.535911602209945,
|
| 836 |
+
"grad_norm": 0.09005668759346008,
|
| 837 |
+
"learning_rate": 7.188459177409454e-06,
|
| 838 |
+
"loss": 0.1148,
|
| 839 |
+
"step": 2450
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 13.67403314917127,
|
| 843 |
+
"grad_norm": 0.0306901503354311,
|
| 844 |
+
"learning_rate": 7.0349907918968695e-06,
|
| 845 |
+
"loss": 0.0373,
|
| 846 |
+
"step": 2475
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 13.812154696132596,
|
| 850 |
+
"grad_norm": 0.04027491435408592,
|
| 851 |
+
"learning_rate": 6.8815224063842856e-06,
|
| 852 |
+
"loss": 0.0203,
|
| 853 |
+
"step": 2500
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"epoch": 13.950276243093922,
|
| 857 |
+
"grad_norm": 0.031085703521966934,
|
| 858 |
+
"learning_rate": 6.728054020871701e-06,
|
| 859 |
+
"loss": 0.2028,
|
| 860 |
+
"step": 2525
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 14.0,
|
| 864 |
+
"eval_accuracy": 0.8436426116838488,
|
| 865 |
+
"eval_loss": 1.15190589427948,
|
| 866 |
+
"eval_macro_f1": 0.8429327056126862,
|
| 867 |
+
"eval_runtime": 1.0265,
|
| 868 |
+
"eval_samples_per_second": 566.951,
|
| 869 |
+
"eval_steps_per_second": 36.043,
|
| 870 |
+
"eval_weighted_f1": 0.8433719802479362,
|
| 871 |
+
"step": 2534
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 14.088397790055248,
|
| 875 |
+
"grad_norm": 0.02857493981719017,
|
| 876 |
+
"learning_rate": 6.574585635359117e-06,
|
| 877 |
+
"loss": 0.0049,
|
| 878 |
+
"step": 2550
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 14.226519337016574,
|
| 882 |
+
"grad_norm": 0.02868039719760418,
|
| 883 |
+
"learning_rate": 6.421117249846533e-06,
|
| 884 |
+
"loss": 0.0808,
|
| 885 |
+
"step": 2575
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 14.3646408839779,
|
| 889 |
+
"grad_norm": 1.7899531126022339,
|
| 890 |
+
"learning_rate": 6.267648864333947e-06,
|
| 891 |
+
"loss": 0.0541,
|
| 892 |
+
"step": 2600
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 14.502762430939226,
|
| 896 |
+
"grad_norm": 0.9857421517372131,
|
| 897 |
+
"learning_rate": 6.114180478821363e-06,
|
| 898 |
+
"loss": 0.1246,
|
| 899 |
+
"step": 2625
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 14.640883977900552,
|
| 903 |
+
"grad_norm": 0.029640762135386467,
|
| 904 |
+
"learning_rate": 5.960712093308779e-06,
|
| 905 |
+
"loss": 0.1541,
|
| 906 |
+
"step": 2650
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 14.779005524861878,
|
| 910 |
+
"grad_norm": 0.03865622356534004,
|
| 911 |
+
"learning_rate": 5.807243707796194e-06,
|
| 912 |
+
"loss": 0.0889,
|
| 913 |
+
"step": 2675
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 14.917127071823204,
|
| 917 |
+
"grad_norm": 0.028642505407333374,
|
| 918 |
+
"learning_rate": 5.65377532228361e-06,
|
| 919 |
+
"loss": 0.0552,
|
| 920 |
+
"step": 2700
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 15.0,
|
| 924 |
+
"eval_accuracy": 0.845360824742268,
|
| 925 |
+
"eval_loss": 1.1531429290771484,
|
| 926 |
+
"eval_macro_f1": 0.8450306048345264,
|
| 927 |
+
"eval_runtime": 1.1148,
|
| 928 |
+
"eval_samples_per_second": 522.088,
|
| 929 |
+
"eval_steps_per_second": 33.191,
|
| 930 |
+
"eval_weighted_f1": 0.8454929127053645,
|
| 931 |
+
"step": 2715
|
| 932 |
+
}
|
| 933 |
+
],
|
| 934 |
+
"logging_steps": 25,
|
| 935 |
+
"max_steps": 3620,
|
| 936 |
+
"num_input_tokens_seen": 0,
|
| 937 |
+
"num_train_epochs": 20,
|
| 938 |
+
"save_steps": 500,
|
| 939 |
+
"stateful_callbacks": {
|
| 940 |
+
"EarlyStoppingCallback": {
|
| 941 |
+
"args": {
|
| 942 |
+
"early_stopping_patience": 3,
|
| 943 |
+
"early_stopping_threshold": 0.0
|
| 944 |
+
},
|
| 945 |
+
"attributes": {
|
| 946 |
+
"early_stopping_patience_counter": 0
|
| 947 |
+
}
|
| 948 |
+
},
|
| 949 |
+
"TrainerControl": {
|
| 950 |
+
"args": {
|
| 951 |
+
"should_epoch_stop": false,
|
| 952 |
+
"should_evaluate": false,
|
| 953 |
+
"should_log": false,
|
| 954 |
+
"should_save": true,
|
| 955 |
+
"should_training_stop": false
|
| 956 |
+
},
|
| 957 |
+
"attributes": {}
|
| 958 |
+
}
|
| 959 |
+
},
|
| 960 |
+
"total_flos": 0.0,
|
| 961 |
+
"train_batch_size": 8,
|
| 962 |
+
"trial_name": null,
|
| 963 |
+
"trial_params": null
|
| 964 |
+
}
|
ood-vifactcheck-bamibert-source-source-02-s43-20260829T004338Z-3858bb98/trainer/checkpoint-3258/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-source-source-02-s43-20260829T004338Z-3858bb98/trainer/checkpoint-3258/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-vifactcheck-bamibert-source-source-02-s43-20260829T004338Z-3858bb98/trainer/checkpoint-3258/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-source-source-02-s43-20260829T004338Z-3858bb98/trainer/checkpoint-3258/trainer_state.json
ADDED
|
@@ -0,0 +1,1151 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 2715,
|
| 3 |
+
"best_metric": 0.8450306048345264,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-source-source-02-s43-20260829T004338Z-3858bb98/trainer/checkpoint-2715",
|
| 5 |
+
"epoch": 18.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 3258,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.13812154696132597,
|
| 14 |
+
"grad_norm": 11.385866165161133,
|
| 15 |
+
"learning_rate": 1.3259668508287293e-06,
|
| 16 |
+
"loss": 4.4148,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.27624309392265195,
|
| 21 |
+
"grad_norm": 8.704425811767578,
|
| 22 |
+
"learning_rate": 2.707182320441989e-06,
|
| 23 |
+
"loss": 4.4207,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.4143646408839779,
|
| 28 |
+
"grad_norm": 14.337090492248535,
|
| 29 |
+
"learning_rate": 4.088397790055249e-06,
|
| 30 |
+
"loss": 4.4151,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5524861878453039,
|
| 35 |
+
"grad_norm": 162.95692443847656,
|
| 36 |
+
"learning_rate": 5.469613259668509e-06,
|
| 37 |
+
"loss": 4.3881,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.6906077348066298,
|
| 42 |
+
"grad_norm": 12.436147689819336,
|
| 43 |
+
"learning_rate": 6.850828729281769e-06,
|
| 44 |
+
"loss": 4.3728,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.8287292817679558,
|
| 49 |
+
"grad_norm": 19.559032440185547,
|
| 50 |
+
"learning_rate": 8.232044198895029e-06,
|
| 51 |
+
"loss": 4.3198,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.9668508287292817,
|
| 56 |
+
"grad_norm": 20.87177085876465,
|
| 57 |
+
"learning_rate": 9.613259668508288e-06,
|
| 58 |
+
"loss": 4.3045,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 1.0,
|
| 63 |
+
"eval_accuracy": 0.48625429553264604,
|
| 64 |
+
"eval_loss": 1.0244207382202148,
|
| 65 |
+
"eval_macro_f1": 0.4643918568040181,
|
| 66 |
+
"eval_runtime": 1.1672,
|
| 67 |
+
"eval_samples_per_second": 498.609,
|
| 68 |
+
"eval_steps_per_second": 31.699,
|
| 69 |
+
"eval_weighted_f1": 0.46821140597701794,
|
| 70 |
+
"step": 181
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.1049723756906078,
|
| 74 |
+
"grad_norm": 38.50503921508789,
|
| 75 |
+
"learning_rate": 1.0994475138121547e-05,
|
| 76 |
+
"loss": 4.0763,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.2430939226519337,
|
| 81 |
+
"grad_norm": 32.16830062866211,
|
| 82 |
+
"learning_rate": 1.2375690607734806e-05,
|
| 83 |
+
"loss": 3.6891,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.3812154696132597,
|
| 88 |
+
"grad_norm": 57.03675842285156,
|
| 89 |
+
"learning_rate": 1.3756906077348067e-05,
|
| 90 |
+
"loss": 3.3411,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.5193370165745856,
|
| 95 |
+
"grad_norm": 59.425662994384766,
|
| 96 |
+
"learning_rate": 1.5138121546961326e-05,
|
| 97 |
+
"loss": 3.0682,
|
| 98 |
+
"step": 275
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.6574585635359116,
|
| 102 |
+
"grad_norm": 38.729461669921875,
|
| 103 |
+
"learning_rate": 1.6519337016574586e-05,
|
| 104 |
+
"loss": 2.9975,
|
| 105 |
+
"step": 300
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.7955801104972375,
|
| 109 |
+
"grad_norm": 65.91519165039062,
|
| 110 |
+
"learning_rate": 1.7900552486187846e-05,
|
| 111 |
+
"loss": 2.9397,
|
| 112 |
+
"step": 325
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 1.9337016574585635,
|
| 116 |
+
"grad_norm": 57.606727600097656,
|
| 117 |
+
"learning_rate": 1.9281767955801107e-05,
|
| 118 |
+
"loss": 2.5799,
|
| 119 |
+
"step": 350
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 2.0,
|
| 123 |
+
"eval_accuracy": 0.7422680412371134,
|
| 124 |
+
"eval_loss": 0.6076148152351379,
|
| 125 |
+
"eval_macro_f1": 0.729084090886048,
|
| 126 |
+
"eval_runtime": 1.138,
|
| 127 |
+
"eval_samples_per_second": 511.433,
|
| 128 |
+
"eval_steps_per_second": 32.514,
|
| 129 |
+
"eval_weighted_f1": 0.7309228992577754,
|
| 130 |
+
"step": 362
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.0718232044198897,
|
| 134 |
+
"grad_norm": 65.74261474609375,
|
| 135 |
+
"learning_rate": 1.9926335174953962e-05,
|
| 136 |
+
"loss": 2.3743,
|
| 137 |
+
"step": 375
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 2.2099447513812156,
|
| 141 |
+
"grad_norm": 66.28155517578125,
|
| 142 |
+
"learning_rate": 1.9772866789441377e-05,
|
| 143 |
+
"loss": 2.2332,
|
| 144 |
+
"step": 400
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.3480662983425415,
|
| 148 |
+
"grad_norm": 60.59415817260742,
|
| 149 |
+
"learning_rate": 1.9619398403928794e-05,
|
| 150 |
+
"loss": 2.2188,
|
| 151 |
+
"step": 425
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.4861878453038675,
|
| 155 |
+
"grad_norm": 80.37751770019531,
|
| 156 |
+
"learning_rate": 1.946593001841621e-05,
|
| 157 |
+
"loss": 2.1052,
|
| 158 |
+
"step": 450
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.6243093922651934,
|
| 162 |
+
"grad_norm": 53.990535736083984,
|
| 163 |
+
"learning_rate": 1.9312461632903623e-05,
|
| 164 |
+
"loss": 2.0804,
|
| 165 |
+
"step": 475
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.7624309392265194,
|
| 169 |
+
"grad_norm": 38.29616928100586,
|
| 170 |
+
"learning_rate": 1.915899324739104e-05,
|
| 171 |
+
"loss": 1.9099,
|
| 172 |
+
"step": 500
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 2.9005524861878453,
|
| 176 |
+
"grad_norm": 26.27406883239746,
|
| 177 |
+
"learning_rate": 1.9005524861878455e-05,
|
| 178 |
+
"loss": 1.9504,
|
| 179 |
+
"step": 525
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 3.0,
|
| 183 |
+
"eval_accuracy": 0.7800687285223368,
|
| 184 |
+
"eval_loss": 0.5324428081512451,
|
| 185 |
+
"eval_macro_f1": 0.7815604168357787,
|
| 186 |
+
"eval_runtime": 1.1299,
|
| 187 |
+
"eval_samples_per_second": 515.079,
|
| 188 |
+
"eval_steps_per_second": 32.746,
|
| 189 |
+
"eval_weighted_f1": 0.7826158434225067,
|
| 190 |
+
"step": 543
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 3.0386740331491713,
|
| 194 |
+
"grad_norm": 36.635494232177734,
|
| 195 |
+
"learning_rate": 1.885205647636587e-05,
|
| 196 |
+
"loss": 1.9231,
|
| 197 |
+
"step": 550
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 3.1767955801104972,
|
| 201 |
+
"grad_norm": 38.11379623413086,
|
| 202 |
+
"learning_rate": 1.8698588090853287e-05,
|
| 203 |
+
"loss": 1.4116,
|
| 204 |
+
"step": 575
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.314917127071823,
|
| 208 |
+
"grad_norm": 31.0555477142334,
|
| 209 |
+
"learning_rate": 1.85451197053407e-05,
|
| 210 |
+
"loss": 1.7786,
|
| 211 |
+
"step": 600
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.453038674033149,
|
| 215 |
+
"grad_norm": 62.93512725830078,
|
| 216 |
+
"learning_rate": 1.8391651319828116e-05,
|
| 217 |
+
"loss": 1.462,
|
| 218 |
+
"step": 625
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.591160220994475,
|
| 222 |
+
"grad_norm": 44.05408477783203,
|
| 223 |
+
"learning_rate": 1.823818293431553e-05,
|
| 224 |
+
"loss": 1.5503,
|
| 225 |
+
"step": 650
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.729281767955801,
|
| 229 |
+
"grad_norm": 27.487030029296875,
|
| 230 |
+
"learning_rate": 1.8084714548802948e-05,
|
| 231 |
+
"loss": 1.399,
|
| 232 |
+
"step": 675
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.867403314917127,
|
| 236 |
+
"grad_norm": 41.88737487792969,
|
| 237 |
+
"learning_rate": 1.7931246163290362e-05,
|
| 238 |
+
"loss": 1.4536,
|
| 239 |
+
"step": 700
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 4.0,
|
| 243 |
+
"eval_accuracy": 0.7955326460481099,
|
| 244 |
+
"eval_loss": 0.591875433921814,
|
| 245 |
+
"eval_macro_f1": 0.795489967546203,
|
| 246 |
+
"eval_runtime": 1.0189,
|
| 247 |
+
"eval_samples_per_second": 571.178,
|
| 248 |
+
"eval_steps_per_second": 36.312,
|
| 249 |
+
"eval_weighted_f1": 0.7956920276233653,
|
| 250 |
+
"step": 724
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 4.005524861878453,
|
| 254 |
+
"grad_norm": 40.481544494628906,
|
| 255 |
+
"learning_rate": 1.7777777777777777e-05,
|
| 256 |
+
"loss": 1.5514,
|
| 257 |
+
"step": 725
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 4.143646408839779,
|
| 261 |
+
"grad_norm": 62.872493743896484,
|
| 262 |
+
"learning_rate": 1.7624309392265194e-05,
|
| 263 |
+
"loss": 0.8882,
|
| 264 |
+
"step": 750
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.281767955801105,
|
| 268 |
+
"grad_norm": 31.99635887145996,
|
| 269 |
+
"learning_rate": 1.747084100675261e-05,
|
| 270 |
+
"loss": 0.917,
|
| 271 |
+
"step": 775
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 4.419889502762431,
|
| 275 |
+
"grad_norm": 95.87877655029297,
|
| 276 |
+
"learning_rate": 1.7317372621240026e-05,
|
| 277 |
+
"loss": 1.0126,
|
| 278 |
+
"step": 800
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 4.558011049723757,
|
| 282 |
+
"grad_norm": 20.457937240600586,
|
| 283 |
+
"learning_rate": 1.716390423572744e-05,
|
| 284 |
+
"loss": 0.9183,
|
| 285 |
+
"step": 825
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 4.696132596685083,
|
| 289 |
+
"grad_norm": 69.46328735351562,
|
| 290 |
+
"learning_rate": 1.701043585021486e-05,
|
| 291 |
+
"loss": 0.8732,
|
| 292 |
+
"step": 850
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 4.834254143646409,
|
| 296 |
+
"grad_norm": 43.254112243652344,
|
| 297 |
+
"learning_rate": 1.6856967464702273e-05,
|
| 298 |
+
"loss": 0.9655,
|
| 299 |
+
"step": 875
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 4.972375690607735,
|
| 303 |
+
"grad_norm": 57.20595932006836,
|
| 304 |
+
"learning_rate": 1.670349907918969e-05,
|
| 305 |
+
"loss": 1.0731,
|
| 306 |
+
"step": 900
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 5.0,
|
| 310 |
+
"eval_accuracy": 0.802405498281787,
|
| 311 |
+
"eval_loss": 0.5767754316329956,
|
| 312 |
+
"eval_macro_f1": 0.8012312938988065,
|
| 313 |
+
"eval_runtime": 1.1082,
|
| 314 |
+
"eval_samples_per_second": 525.19,
|
| 315 |
+
"eval_steps_per_second": 33.388,
|
| 316 |
+
"eval_weighted_f1": 0.8020146931073064,
|
| 317 |
+
"step": 905
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 5.110497237569061,
|
| 321 |
+
"grad_norm": 30.346569061279297,
|
| 322 |
+
"learning_rate": 1.6550030693677105e-05,
|
| 323 |
+
"loss": 0.6757,
|
| 324 |
+
"step": 925
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 5.248618784530387,
|
| 328 |
+
"grad_norm": 79.27243041992188,
|
| 329 |
+
"learning_rate": 1.639656230816452e-05,
|
| 330 |
+
"loss": 0.7229,
|
| 331 |
+
"step": 950
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.386740331491713,
|
| 335 |
+
"grad_norm": 76.34603118896484,
|
| 336 |
+
"learning_rate": 1.6243093922651937e-05,
|
| 337 |
+
"loss": 0.7669,
|
| 338 |
+
"step": 975
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 5.524861878453039,
|
| 342 |
+
"grad_norm": 39.100955963134766,
|
| 343 |
+
"learning_rate": 1.608962553713935e-05,
|
| 344 |
+
"loss": 0.7318,
|
| 345 |
+
"step": 1000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.662983425414365,
|
| 349 |
+
"grad_norm": 44.392601013183594,
|
| 350 |
+
"learning_rate": 1.5936157151626766e-05,
|
| 351 |
+
"loss": 0.6025,
|
| 352 |
+
"step": 1025
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 5.801104972375691,
|
| 356 |
+
"grad_norm": 76.7225112915039,
|
| 357 |
+
"learning_rate": 1.5782688766114183e-05,
|
| 358 |
+
"loss": 0.6433,
|
| 359 |
+
"step": 1050
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 5.939226519337017,
|
| 363 |
+
"grad_norm": 26.3995361328125,
|
| 364 |
+
"learning_rate": 1.5629220380601598e-05,
|
| 365 |
+
"loss": 0.7356,
|
| 366 |
+
"step": 1075
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 6.0,
|
| 370 |
+
"eval_accuracy": 0.8041237113402062,
|
| 371 |
+
"eval_loss": 0.755528450012207,
|
| 372 |
+
"eval_macro_f1": 0.8035958989131342,
|
| 373 |
+
"eval_runtime": 1.1151,
|
| 374 |
+
"eval_samples_per_second": 521.931,
|
| 375 |
+
"eval_steps_per_second": 33.181,
|
| 376 |
+
"eval_weighted_f1": 0.8039718881665117,
|
| 377 |
+
"step": 1086
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 6.077348066298343,
|
| 381 |
+
"grad_norm": 16.3472900390625,
|
| 382 |
+
"learning_rate": 1.5475751995089012e-05,
|
| 383 |
+
"loss": 0.7126,
|
| 384 |
+
"step": 1100
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 6.2154696132596685,
|
| 388 |
+
"grad_norm": 67.30068969726562,
|
| 389 |
+
"learning_rate": 1.532228360957643e-05,
|
| 390 |
+
"loss": 0.4377,
|
| 391 |
+
"step": 1125
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 6.3535911602209945,
|
| 395 |
+
"grad_norm": 15.90624713897705,
|
| 396 |
+
"learning_rate": 1.5168815224063844e-05,
|
| 397 |
+
"loss": 0.4894,
|
| 398 |
+
"step": 1150
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 6.49171270718232,
|
| 402 |
+
"grad_norm": 119.05094146728516,
|
| 403 |
+
"learning_rate": 1.5015346838551258e-05,
|
| 404 |
+
"loss": 0.5446,
|
| 405 |
+
"step": 1175
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 6.629834254143646,
|
| 409 |
+
"grad_norm": 6.7117156982421875,
|
| 410 |
+
"learning_rate": 1.4861878453038676e-05,
|
| 411 |
+
"loss": 0.4783,
|
| 412 |
+
"step": 1200
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 6.767955801104972,
|
| 416 |
+
"grad_norm": 81.19427490234375,
|
| 417 |
+
"learning_rate": 1.470841006752609e-05,
|
| 418 |
+
"loss": 0.5304,
|
| 419 |
+
"step": 1225
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 6.906077348066298,
|
| 423 |
+
"grad_norm": 71.07677459716797,
|
| 424 |
+
"learning_rate": 1.4554941682013506e-05,
|
| 425 |
+
"loss": 0.4676,
|
| 426 |
+
"step": 1250
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 7.0,
|
| 430 |
+
"eval_accuracy": 0.8127147766323024,
|
| 431 |
+
"eval_loss": 0.8740180134773254,
|
| 432 |
+
"eval_macro_f1": 0.8119854921936865,
|
| 433 |
+
"eval_runtime": 1.0605,
|
| 434 |
+
"eval_samples_per_second": 548.792,
|
| 435 |
+
"eval_steps_per_second": 34.889,
|
| 436 |
+
"eval_weighted_f1": 0.8123917469310626,
|
| 437 |
+
"step": 1267
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 7.044198895027624,
|
| 441 |
+
"grad_norm": 3.1372318267822266,
|
| 442 |
+
"learning_rate": 1.4401473296500922e-05,
|
| 443 |
+
"loss": 0.4766,
|
| 444 |
+
"step": 1275
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 7.18232044198895,
|
| 448 |
+
"grad_norm": 24.158937454223633,
|
| 449 |
+
"learning_rate": 1.4248004910988338e-05,
|
| 450 |
+
"loss": 0.3425,
|
| 451 |
+
"step": 1300
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 7.320441988950276,
|
| 455 |
+
"grad_norm": 40.12568664550781,
|
| 456 |
+
"learning_rate": 1.4094536525475753e-05,
|
| 457 |
+
"loss": 0.3982,
|
| 458 |
+
"step": 1325
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 7.458563535911602,
|
| 462 |
+
"grad_norm": 105.64920043945312,
|
| 463 |
+
"learning_rate": 1.394106813996317e-05,
|
| 464 |
+
"loss": 0.4014,
|
| 465 |
+
"step": 1350
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 7.596685082872928,
|
| 469 |
+
"grad_norm": 17.37604522705078,
|
| 470 |
+
"learning_rate": 1.3787599754450585e-05,
|
| 471 |
+
"loss": 0.4638,
|
| 472 |
+
"step": 1375
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 7.734806629834254,
|
| 476 |
+
"grad_norm": 113.165283203125,
|
| 477 |
+
"learning_rate": 1.3634131368937999e-05,
|
| 478 |
+
"loss": 0.3836,
|
| 479 |
+
"step": 1400
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 7.87292817679558,
|
| 483 |
+
"grad_norm": 49.570980072021484,
|
| 484 |
+
"learning_rate": 1.3480662983425417e-05,
|
| 485 |
+
"loss": 0.434,
|
| 486 |
+
"step": 1425
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 8.0,
|
| 490 |
+
"eval_accuracy": 0.8075601374570447,
|
| 491 |
+
"eval_loss": 1.010019302368164,
|
| 492 |
+
"eval_macro_f1": 0.8053691385981757,
|
| 493 |
+
"eval_runtime": 1.0948,
|
| 494 |
+
"eval_samples_per_second": 531.611,
|
| 495 |
+
"eval_steps_per_second": 33.797,
|
| 496 |
+
"eval_weighted_f1": 0.8064679133325544,
|
| 497 |
+
"step": 1448
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 8.011049723756907,
|
| 501 |
+
"grad_norm": 41.95655059814453,
|
| 502 |
+
"learning_rate": 1.3327194597912831e-05,
|
| 503 |
+
"loss": 0.4402,
|
| 504 |
+
"step": 1450
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 8.149171270718233,
|
| 508 |
+
"grad_norm": 37.319374084472656,
|
| 509 |
+
"learning_rate": 1.3173726212400245e-05,
|
| 510 |
+
"loss": 0.192,
|
| 511 |
+
"step": 1475
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 8.287292817679559,
|
| 515 |
+
"grad_norm": 5.263917922973633,
|
| 516 |
+
"learning_rate": 1.3020257826887662e-05,
|
| 517 |
+
"loss": 0.3384,
|
| 518 |
+
"step": 1500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 8.425414364640885,
|
| 522 |
+
"grad_norm": 4.049027919769287,
|
| 523 |
+
"learning_rate": 1.2866789441375078e-05,
|
| 524 |
+
"loss": 0.3778,
|
| 525 |
+
"step": 1525
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 8.56353591160221,
|
| 529 |
+
"grad_norm": 134.29124450683594,
|
| 530 |
+
"learning_rate": 1.2713321055862494e-05,
|
| 531 |
+
"loss": 0.3379,
|
| 532 |
+
"step": 1550
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 8.701657458563536,
|
| 536 |
+
"grad_norm": 118.86800384521484,
|
| 537 |
+
"learning_rate": 1.2559852670349908e-05,
|
| 538 |
+
"loss": 0.2676,
|
| 539 |
+
"step": 1575
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 8.839779005524862,
|
| 543 |
+
"grad_norm": 0.21576865017414093,
|
| 544 |
+
"learning_rate": 1.2406384284837326e-05,
|
| 545 |
+
"loss": 0.1817,
|
| 546 |
+
"step": 1600
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 8.977900552486188,
|
| 550 |
+
"grad_norm": 266.0789794921875,
|
| 551 |
+
"learning_rate": 1.225291589932474e-05,
|
| 552 |
+
"loss": 0.437,
|
| 553 |
+
"step": 1625
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 9.0,
|
| 557 |
+
"eval_accuracy": 0.8230240549828178,
|
| 558 |
+
"eval_loss": 1.01772940158844,
|
| 559 |
+
"eval_macro_f1": 0.8219837928724036,
|
| 560 |
+
"eval_runtime": 1.0465,
|
| 561 |
+
"eval_samples_per_second": 556.148,
|
| 562 |
+
"eval_steps_per_second": 35.356,
|
| 563 |
+
"eval_weighted_f1": 0.8225489622623926,
|
| 564 |
+
"step": 1629
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 9.116022099447514,
|
| 568 |
+
"grad_norm": 6.092472076416016,
|
| 569 |
+
"learning_rate": 1.2099447513812154e-05,
|
| 570 |
+
"loss": 0.2903,
|
| 571 |
+
"step": 1650
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 9.25414364640884,
|
| 575 |
+
"grad_norm": 3.596750497817993,
|
| 576 |
+
"learning_rate": 1.1945979128299572e-05,
|
| 577 |
+
"loss": 0.2076,
|
| 578 |
+
"step": 1675
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 9.392265193370166,
|
| 582 |
+
"grad_norm": 2.6152288913726807,
|
| 583 |
+
"learning_rate": 1.1792510742786986e-05,
|
| 584 |
+
"loss": 0.3777,
|
| 585 |
+
"step": 1700
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 9.530386740331492,
|
| 589 |
+
"grad_norm": 5.998469352722168,
|
| 590 |
+
"learning_rate": 1.1639042357274402e-05,
|
| 591 |
+
"loss": 0.2567,
|
| 592 |
+
"step": 1725
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 9.668508287292818,
|
| 596 |
+
"grad_norm": 63.65530776977539,
|
| 597 |
+
"learning_rate": 1.1485573971761818e-05,
|
| 598 |
+
"loss": 0.3461,
|
| 599 |
+
"step": 1750
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 9.806629834254144,
|
| 603 |
+
"grad_norm": 289.3068542480469,
|
| 604 |
+
"learning_rate": 1.1332105586249233e-05,
|
| 605 |
+
"loss": 0.2654,
|
| 606 |
+
"step": 1775
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 9.94475138121547,
|
| 610 |
+
"grad_norm": 0.29069459438323975,
|
| 611 |
+
"learning_rate": 1.1178637200736649e-05,
|
| 612 |
+
"loss": 0.2365,
|
| 613 |
+
"step": 1800
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 10.0,
|
| 617 |
+
"eval_accuracy": 0.8230240549828178,
|
| 618 |
+
"eval_loss": 1.0735164880752563,
|
| 619 |
+
"eval_macro_f1": 0.8211941595494235,
|
| 620 |
+
"eval_runtime": 1.1903,
|
| 621 |
+
"eval_samples_per_second": 488.952,
|
| 622 |
+
"eval_steps_per_second": 31.085,
|
| 623 |
+
"eval_weighted_f1": 0.822216029780875,
|
| 624 |
+
"step": 1810
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 10.082872928176796,
|
| 628 |
+
"grad_norm": 0.36485153436660767,
|
| 629 |
+
"learning_rate": 1.1025168815224065e-05,
|
| 630 |
+
"loss": 0.1739,
|
| 631 |
+
"step": 1825
|
| 632 |
+
},
|
| 633 |
+
{
|
| 634 |
+
"epoch": 10.220994475138122,
|
| 635 |
+
"grad_norm": 0.6285772323608398,
|
| 636 |
+
"learning_rate": 1.087170042971148e-05,
|
| 637 |
+
"loss": 0.1638,
|
| 638 |
+
"step": 1850
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 10.359116022099448,
|
| 642 |
+
"grad_norm": 212.32395935058594,
|
| 643 |
+
"learning_rate": 1.0718232044198895e-05,
|
| 644 |
+
"loss": 0.2131,
|
| 645 |
+
"step": 1875
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 10.497237569060774,
|
| 649 |
+
"grad_norm": 265.7480773925781,
|
| 650 |
+
"learning_rate": 1.0564763658686313e-05,
|
| 651 |
+
"loss": 0.2143,
|
| 652 |
+
"step": 1900
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"epoch": 10.6353591160221,
|
| 656 |
+
"grad_norm": 0.04942381754517555,
|
| 657 |
+
"learning_rate": 1.0411295273173727e-05,
|
| 658 |
+
"loss": 0.1551,
|
| 659 |
+
"step": 1925
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 10.773480662983426,
|
| 663 |
+
"grad_norm": 39.93251037597656,
|
| 664 |
+
"learning_rate": 1.0257826887661141e-05,
|
| 665 |
+
"loss": 0.1928,
|
| 666 |
+
"step": 1950
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 10.911602209944752,
|
| 670 |
+
"grad_norm": 0.06080511212348938,
|
| 671 |
+
"learning_rate": 1.010435850214856e-05,
|
| 672 |
+
"loss": 0.1907,
|
| 673 |
+
"step": 1975
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 11.0,
|
| 677 |
+
"eval_accuracy": 0.8264604810996563,
|
| 678 |
+
"eval_loss": 1.064011573791504,
|
| 679 |
+
"eval_macro_f1": 0.8259139762802074,
|
| 680 |
+
"eval_runtime": 1.015,
|
| 681 |
+
"eval_samples_per_second": 573.375,
|
| 682 |
+
"eval_steps_per_second": 36.452,
|
| 683 |
+
"eval_weighted_f1": 0.8267390565929533,
|
| 684 |
+
"step": 1991
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 11.049723756906078,
|
| 688 |
+
"grad_norm": 1.0326106548309326,
|
| 689 |
+
"learning_rate": 9.950890116635974e-06,
|
| 690 |
+
"loss": 0.1865,
|
| 691 |
+
"step": 2000
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 11.187845303867404,
|
| 695 |
+
"grad_norm": 149.4682159423828,
|
| 696 |
+
"learning_rate": 9.79742173112339e-06,
|
| 697 |
+
"loss": 0.1998,
|
| 698 |
+
"step": 2025
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 11.32596685082873,
|
| 702 |
+
"grad_norm": 0.06841234862804413,
|
| 703 |
+
"learning_rate": 9.643953345610806e-06,
|
| 704 |
+
"loss": 0.1445,
|
| 705 |
+
"step": 2050
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 11.464088397790055,
|
| 709 |
+
"grad_norm": 25.998315811157227,
|
| 710 |
+
"learning_rate": 9.49048496009822e-06,
|
| 711 |
+
"loss": 0.1885,
|
| 712 |
+
"step": 2075
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 11.602209944751381,
|
| 716 |
+
"grad_norm": 0.6273815035820007,
|
| 717 |
+
"learning_rate": 9.337016574585636e-06,
|
| 718 |
+
"loss": 0.2335,
|
| 719 |
+
"step": 2100
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 11.740331491712707,
|
| 723 |
+
"grad_norm": 0.15333878993988037,
|
| 724 |
+
"learning_rate": 9.183548189073052e-06,
|
| 725 |
+
"loss": 0.1139,
|
| 726 |
+
"step": 2125
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 11.878453038674033,
|
| 730 |
+
"grad_norm": 0.04380929842591286,
|
| 731 |
+
"learning_rate": 9.030079803560468e-06,
|
| 732 |
+
"loss": 0.1818,
|
| 733 |
+
"step": 2150
|
| 734 |
+
},
|
| 735 |
+
{
|
| 736 |
+
"epoch": 12.0,
|
| 737 |
+
"eval_accuracy": 0.8281786941580757,
|
| 738 |
+
"eval_loss": 1.1222890615463257,
|
| 739 |
+
"eval_macro_f1": 0.8271145219966799,
|
| 740 |
+
"eval_runtime": 1.1021,
|
| 741 |
+
"eval_samples_per_second": 528.097,
|
| 742 |
+
"eval_steps_per_second": 33.573,
|
| 743 |
+
"eval_weighted_f1": 0.8276554250282702,
|
| 744 |
+
"step": 2172
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 12.01657458563536,
|
| 748 |
+
"grad_norm": 0.09283286333084106,
|
| 749 |
+
"learning_rate": 8.876611418047882e-06,
|
| 750 |
+
"loss": 0.2025,
|
| 751 |
+
"step": 2175
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 12.154696132596685,
|
| 755 |
+
"grad_norm": 0.03827335685491562,
|
| 756 |
+
"learning_rate": 8.723143032535298e-06,
|
| 757 |
+
"loss": 0.0699,
|
| 758 |
+
"step": 2200
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 12.292817679558011,
|
| 762 |
+
"grad_norm": 0.24005241692066193,
|
| 763 |
+
"learning_rate": 8.569674647022714e-06,
|
| 764 |
+
"loss": 0.1297,
|
| 765 |
+
"step": 2225
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 12.430939226519337,
|
| 769 |
+
"grad_norm": 3.1738510131835938,
|
| 770 |
+
"learning_rate": 8.416206261510129e-06,
|
| 771 |
+
"loss": 0.0815,
|
| 772 |
+
"step": 2250
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 12.569060773480663,
|
| 776 |
+
"grad_norm": 25.839738845825195,
|
| 777 |
+
"learning_rate": 8.262737875997545e-06,
|
| 778 |
+
"loss": 0.1777,
|
| 779 |
+
"step": 2275
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 12.707182320441989,
|
| 783 |
+
"grad_norm": 0.046633850783109665,
|
| 784 |
+
"learning_rate": 8.10926949048496e-06,
|
| 785 |
+
"loss": 0.0432,
|
| 786 |
+
"step": 2300
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 12.845303867403315,
|
| 790 |
+
"grad_norm": 486.20989990234375,
|
| 791 |
+
"learning_rate": 7.955801104972377e-06,
|
| 792 |
+
"loss": 0.0231,
|
| 793 |
+
"step": 2325
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 12.98342541436464,
|
| 797 |
+
"grad_norm": 0.04107821732759476,
|
| 798 |
+
"learning_rate": 7.802332719459793e-06,
|
| 799 |
+
"loss": 0.1274,
|
| 800 |
+
"step": 2350
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 13.0,
|
| 804 |
+
"eval_accuracy": 0.8298969072164949,
|
| 805 |
+
"eval_loss": 1.1414563655853271,
|
| 806 |
+
"eval_macro_f1": 0.8294064887271121,
|
| 807 |
+
"eval_runtime": 1.2772,
|
| 808 |
+
"eval_samples_per_second": 455.682,
|
| 809 |
+
"eval_steps_per_second": 28.969,
|
| 810 |
+
"eval_weighted_f1": 0.8298931070542709,
|
| 811 |
+
"step": 2353
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"epoch": 13.121546961325967,
|
| 815 |
+
"grad_norm": 354.08575439453125,
|
| 816 |
+
"learning_rate": 7.648864333947207e-06,
|
| 817 |
+
"loss": 0.1001,
|
| 818 |
+
"step": 2375
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 13.259668508287293,
|
| 822 |
+
"grad_norm": 0.16044709086418152,
|
| 823 |
+
"learning_rate": 7.495395948434623e-06,
|
| 824 |
+
"loss": 0.0414,
|
| 825 |
+
"step": 2400
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 13.397790055248619,
|
| 829 |
+
"grad_norm": 0.04437292739748955,
|
| 830 |
+
"learning_rate": 7.341927562922039e-06,
|
| 831 |
+
"loss": 0.0713,
|
| 832 |
+
"step": 2425
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 13.535911602209945,
|
| 836 |
+
"grad_norm": 0.09005668759346008,
|
| 837 |
+
"learning_rate": 7.188459177409454e-06,
|
| 838 |
+
"loss": 0.1148,
|
| 839 |
+
"step": 2450
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 13.67403314917127,
|
| 843 |
+
"grad_norm": 0.0306901503354311,
|
| 844 |
+
"learning_rate": 7.0349907918968695e-06,
|
| 845 |
+
"loss": 0.0373,
|
| 846 |
+
"step": 2475
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 13.812154696132596,
|
| 850 |
+
"grad_norm": 0.04027491435408592,
|
| 851 |
+
"learning_rate": 6.8815224063842856e-06,
|
| 852 |
+
"loss": 0.0203,
|
| 853 |
+
"step": 2500
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"epoch": 13.950276243093922,
|
| 857 |
+
"grad_norm": 0.031085703521966934,
|
| 858 |
+
"learning_rate": 6.728054020871701e-06,
|
| 859 |
+
"loss": 0.2028,
|
| 860 |
+
"step": 2525
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 14.0,
|
| 864 |
+
"eval_accuracy": 0.8436426116838488,
|
| 865 |
+
"eval_loss": 1.15190589427948,
|
| 866 |
+
"eval_macro_f1": 0.8429327056126862,
|
| 867 |
+
"eval_runtime": 1.0265,
|
| 868 |
+
"eval_samples_per_second": 566.951,
|
| 869 |
+
"eval_steps_per_second": 36.043,
|
| 870 |
+
"eval_weighted_f1": 0.8433719802479362,
|
| 871 |
+
"step": 2534
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 14.088397790055248,
|
| 875 |
+
"grad_norm": 0.02857493981719017,
|
| 876 |
+
"learning_rate": 6.574585635359117e-06,
|
| 877 |
+
"loss": 0.0049,
|
| 878 |
+
"step": 2550
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 14.226519337016574,
|
| 882 |
+
"grad_norm": 0.02868039719760418,
|
| 883 |
+
"learning_rate": 6.421117249846533e-06,
|
| 884 |
+
"loss": 0.0808,
|
| 885 |
+
"step": 2575
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 14.3646408839779,
|
| 889 |
+
"grad_norm": 1.7899531126022339,
|
| 890 |
+
"learning_rate": 6.267648864333947e-06,
|
| 891 |
+
"loss": 0.0541,
|
| 892 |
+
"step": 2600
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 14.502762430939226,
|
| 896 |
+
"grad_norm": 0.9857421517372131,
|
| 897 |
+
"learning_rate": 6.114180478821363e-06,
|
| 898 |
+
"loss": 0.1246,
|
| 899 |
+
"step": 2625
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 14.640883977900552,
|
| 903 |
+
"grad_norm": 0.029640762135386467,
|
| 904 |
+
"learning_rate": 5.960712093308779e-06,
|
| 905 |
+
"loss": 0.1541,
|
| 906 |
+
"step": 2650
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 14.779005524861878,
|
| 910 |
+
"grad_norm": 0.03865622356534004,
|
| 911 |
+
"learning_rate": 5.807243707796194e-06,
|
| 912 |
+
"loss": 0.0889,
|
| 913 |
+
"step": 2675
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 14.917127071823204,
|
| 917 |
+
"grad_norm": 0.028642505407333374,
|
| 918 |
+
"learning_rate": 5.65377532228361e-06,
|
| 919 |
+
"loss": 0.0552,
|
| 920 |
+
"step": 2700
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 15.0,
|
| 924 |
+
"eval_accuracy": 0.845360824742268,
|
| 925 |
+
"eval_loss": 1.1531429290771484,
|
| 926 |
+
"eval_macro_f1": 0.8450306048345264,
|
| 927 |
+
"eval_runtime": 1.1148,
|
| 928 |
+
"eval_samples_per_second": 522.088,
|
| 929 |
+
"eval_steps_per_second": 33.191,
|
| 930 |
+
"eval_weighted_f1": 0.8454929127053645,
|
| 931 |
+
"step": 2715
|
| 932 |
+
},
|
| 933 |
+
{
|
| 934 |
+
"epoch": 15.05524861878453,
|
| 935 |
+
"grad_norm": 0.025730909779667854,
|
| 936 |
+
"learning_rate": 5.500306936771026e-06,
|
| 937 |
+
"loss": 0.0977,
|
| 938 |
+
"step": 2725
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 15.193370165745856,
|
| 942 |
+
"grad_norm": 0.4509689211845398,
|
| 943 |
+
"learning_rate": 5.346838551258441e-06,
|
| 944 |
+
"loss": 0.0044,
|
| 945 |
+
"step": 2750
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 15.331491712707182,
|
| 949 |
+
"grad_norm": 0.023593585938215256,
|
| 950 |
+
"learning_rate": 5.193370165745857e-06,
|
| 951 |
+
"loss": 0.099,
|
| 952 |
+
"step": 2775
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"epoch": 15.469613259668508,
|
| 956 |
+
"grad_norm": 0.023794541135430336,
|
| 957 |
+
"learning_rate": 5.039901780233273e-06,
|
| 958 |
+
"loss": 0.0363,
|
| 959 |
+
"step": 2800
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 15.607734806629834,
|
| 963 |
+
"grad_norm": 200.6015625,
|
| 964 |
+
"learning_rate": 4.886433394720688e-06,
|
| 965 |
+
"loss": 0.0527,
|
| 966 |
+
"step": 2825
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 15.74585635359116,
|
| 970 |
+
"grad_norm": 0.07499513775110245,
|
| 971 |
+
"learning_rate": 4.732965009208104e-06,
|
| 972 |
+
"loss": 0.0259,
|
| 973 |
+
"step": 2850
|
| 974 |
+
},
|
| 975 |
+
{
|
| 976 |
+
"epoch": 15.883977900552486,
|
| 977 |
+
"grad_norm": 2.4968767166137695,
|
| 978 |
+
"learning_rate": 4.579496623695519e-06,
|
| 979 |
+
"loss": 0.0591,
|
| 980 |
+
"step": 2875
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 16.0,
|
| 984 |
+
"eval_accuracy": 0.8367697594501718,
|
| 985 |
+
"eval_loss": 1.2312051057815552,
|
| 986 |
+
"eval_macro_f1": 0.8362555112654122,
|
| 987 |
+
"eval_runtime": 0.961,
|
| 988 |
+
"eval_samples_per_second": 605.589,
|
| 989 |
+
"eval_steps_per_second": 38.5,
|
| 990 |
+
"eval_weighted_f1": 0.8368050342589487,
|
| 991 |
+
"step": 2896
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 16.022099447513813,
|
| 995 |
+
"grad_norm": 0.02334602363407612,
|
| 996 |
+
"learning_rate": 4.426028238182934e-06,
|
| 997 |
+
"loss": 0.0395,
|
| 998 |
+
"step": 2900
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 16.16022099447514,
|
| 1002 |
+
"grad_norm": 208.02859497070312,
|
| 1003 |
+
"learning_rate": 4.27255985267035e-06,
|
| 1004 |
+
"loss": 0.0273,
|
| 1005 |
+
"step": 2925
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 16.298342541436465,
|
| 1009 |
+
"grad_norm": 0.024187829345464706,
|
| 1010 |
+
"learning_rate": 4.1190914671577655e-06,
|
| 1011 |
+
"loss": 0.0027,
|
| 1012 |
+
"step": 2950
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 16.43646408839779,
|
| 1016 |
+
"grad_norm": 0.2309839278459549,
|
| 1017 |
+
"learning_rate": 3.9656230816451816e-06,
|
| 1018 |
+
"loss": 0.0421,
|
| 1019 |
+
"step": 2975
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 16.574585635359117,
|
| 1023 |
+
"grad_norm": 0.4096539318561554,
|
| 1024 |
+
"learning_rate": 3.812154696132597e-06,
|
| 1025 |
+
"loss": 0.1715,
|
| 1026 |
+
"step": 3000
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 16.712707182320443,
|
| 1030 |
+
"grad_norm": 0.037988726049661636,
|
| 1031 |
+
"learning_rate": 3.6586863106200128e-06,
|
| 1032 |
+
"loss": 0.0664,
|
| 1033 |
+
"step": 3025
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 16.85082872928177,
|
| 1037 |
+
"grad_norm": 0.02674035355448723,
|
| 1038 |
+
"learning_rate": 3.505217925107428e-06,
|
| 1039 |
+
"loss": 0.0291,
|
| 1040 |
+
"step": 3050
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 16.988950276243095,
|
| 1044 |
+
"grad_norm": 0.021436898037791252,
|
| 1045 |
+
"learning_rate": 3.3517495395948436e-06,
|
| 1046 |
+
"loss": 0.0543,
|
| 1047 |
+
"step": 3075
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 17.0,
|
| 1051 |
+
"eval_accuracy": 0.8281786941580757,
|
| 1052 |
+
"eval_loss": 1.2556906938552856,
|
| 1053 |
+
"eval_macro_f1": 0.8277584372758539,
|
| 1054 |
+
"eval_runtime": 1.0939,
|
| 1055 |
+
"eval_samples_per_second": 532.028,
|
| 1056 |
+
"eval_steps_per_second": 33.823,
|
| 1057 |
+
"eval_weighted_f1": 0.8282389604790302,
|
| 1058 |
+
"step": 3077
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 17.12707182320442,
|
| 1062 |
+
"grad_norm": 0.022764207795262337,
|
| 1063 |
+
"learning_rate": 3.1982811540822596e-06,
|
| 1064 |
+
"loss": 0.1077,
|
| 1065 |
+
"step": 3100
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 17.265193370165747,
|
| 1069 |
+
"grad_norm": 0.19973796606063843,
|
| 1070 |
+
"learning_rate": 3.0448127685696748e-06,
|
| 1071 |
+
"loss": 0.042,
|
| 1072 |
+
"step": 3125
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 17.403314917127073,
|
| 1076 |
+
"grad_norm": 0.023669740185141563,
|
| 1077 |
+
"learning_rate": 2.8913443830570904e-06,
|
| 1078 |
+
"loss": 0.0029,
|
| 1079 |
+
"step": 3150
|
| 1080 |
+
},
|
| 1081 |
+
{
|
| 1082 |
+
"epoch": 17.5414364640884,
|
| 1083 |
+
"grad_norm": 0.022506114095449448,
|
| 1084 |
+
"learning_rate": 2.7378759975445064e-06,
|
| 1085 |
+
"loss": 0.0025,
|
| 1086 |
+
"step": 3175
|
| 1087 |
+
},
|
| 1088 |
+
{
|
| 1089 |
+
"epoch": 17.679558011049725,
|
| 1090 |
+
"grad_norm": 0.031938567757606506,
|
| 1091 |
+
"learning_rate": 2.5844076120319216e-06,
|
| 1092 |
+
"loss": 0.0422,
|
| 1093 |
+
"step": 3200
|
| 1094 |
+
},
|
| 1095 |
+
{
|
| 1096 |
+
"epoch": 17.81767955801105,
|
| 1097 |
+
"grad_norm": 0.024543993175029755,
|
| 1098 |
+
"learning_rate": 2.430939226519337e-06,
|
| 1099 |
+
"loss": 0.0053,
|
| 1100 |
+
"step": 3225
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 17.955801104972377,
|
| 1104 |
+
"grad_norm": 0.02006230503320694,
|
| 1105 |
+
"learning_rate": 2.2774708410067528e-06,
|
| 1106 |
+
"loss": 0.0117,
|
| 1107 |
+
"step": 3250
|
| 1108 |
+
},
|
| 1109 |
+
{
|
| 1110 |
+
"epoch": 18.0,
|
| 1111 |
+
"eval_accuracy": 0.8298969072164949,
|
| 1112 |
+
"eval_loss": 1.2747710943222046,
|
| 1113 |
+
"eval_macro_f1": 0.8295983893922038,
|
| 1114 |
+
"eval_runtime": 1.123,
|
| 1115 |
+
"eval_samples_per_second": 518.253,
|
| 1116 |
+
"eval_steps_per_second": 32.947,
|
| 1117 |
+
"eval_weighted_f1": 0.8300976248473326,
|
| 1118 |
+
"step": 3258
|
| 1119 |
+
}
|
| 1120 |
+
],
|
| 1121 |
+
"logging_steps": 25,
|
| 1122 |
+
"max_steps": 3620,
|
| 1123 |
+
"num_input_tokens_seen": 0,
|
| 1124 |
+
"num_train_epochs": 20,
|
| 1125 |
+
"save_steps": 500,
|
| 1126 |
+
"stateful_callbacks": {
|
| 1127 |
+
"EarlyStoppingCallback": {
|
| 1128 |
+
"args": {
|
| 1129 |
+
"early_stopping_patience": 3,
|
| 1130 |
+
"early_stopping_threshold": 0.0
|
| 1131 |
+
},
|
| 1132 |
+
"attributes": {
|
| 1133 |
+
"early_stopping_patience_counter": 3
|
| 1134 |
+
}
|
| 1135 |
+
},
|
| 1136 |
+
"TrainerControl": {
|
| 1137 |
+
"args": {
|
| 1138 |
+
"should_epoch_stop": false,
|
| 1139 |
+
"should_evaluate": false,
|
| 1140 |
+
"should_log": false,
|
| 1141 |
+
"should_save": true,
|
| 1142 |
+
"should_training_stop": true
|
| 1143 |
+
},
|
| 1144 |
+
"attributes": {}
|
| 1145 |
+
}
|
| 1146 |
+
},
|
| 1147 |
+
"total_flos": 0.0,
|
| 1148 |
+
"train_batch_size": 8,
|
| 1149 |
+
"trial_name": null,
|
| 1150 |
+
"trial_params": null
|
| 1151 |
+
}
|
ood-vifactcheck-bamibert-source-source-03-s42-20260828T232240Z-6235c159/trainer/checkpoint-1464/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-source-source-03-s42-20260828T232240Z-6235c159/trainer/checkpoint-1464/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-vifactcheck-bamibert-source-source-03-s42-20260828T232240Z-6235c159/trainer/checkpoint-1464/trainer_state.json
ADDED
|
@@ -0,0 +1,537 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 915,
|
| 3 |
+
"best_metric": 0.850143080906351,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-source-source-03-s42-20260828T232240Z-6235c159/trainer/checkpoint-915",
|
| 5 |
+
"epoch": 8.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1464,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.13717421124828533,
|
| 14 |
+
"grad_norm": 12.426689147949219,
|
| 15 |
+
"learning_rate": 1.3114754098360657e-06,
|
| 16 |
+
"loss": 4.4639,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.27434842249657065,
|
| 21 |
+
"grad_norm": 40.18903732299805,
|
| 22 |
+
"learning_rate": 2.677595628415301e-06,
|
| 23 |
+
"loss": 4.3942,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.411522633744856,
|
| 28 |
+
"grad_norm": 9.230567932128906,
|
| 29 |
+
"learning_rate": 4.043715846994536e-06,
|
| 30 |
+
"loss": 4.4008,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5486968449931413,
|
| 35 |
+
"grad_norm": 48.2020378112793,
|
| 36 |
+
"learning_rate": 5.409836065573772e-06,
|
| 37 |
+
"loss": 4.3924,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.6858710562414266,
|
| 42 |
+
"grad_norm": 13.302419662475586,
|
| 43 |
+
"learning_rate": 6.775956284153006e-06,
|
| 44 |
+
"loss": 4.3507,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.823045267489712,
|
| 49 |
+
"grad_norm": 15.120257377624512,
|
| 50 |
+
"learning_rate": 8.14207650273224e-06,
|
| 51 |
+
"loss": 4.3756,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.9602194787379973,
|
| 56 |
+
"grad_norm": 20.738765716552734,
|
| 57 |
+
"learning_rate": 9.508196721311476e-06,
|
| 58 |
+
"loss": 4.3333,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 1.0,
|
| 63 |
+
"eval_accuracy": 0.45757071547420963,
|
| 64 |
+
"eval_loss": 1.0564359426498413,
|
| 65 |
+
"eval_macro_f1": 0.42233729113767465,
|
| 66 |
+
"eval_runtime": 1.2782,
|
| 67 |
+
"eval_samples_per_second": 470.182,
|
| 68 |
+
"eval_steps_per_second": 29.729,
|
| 69 |
+
"eval_weighted_f1": 0.426375993368211,
|
| 70 |
+
"step": 183
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.093278463648834,
|
| 74 |
+
"grad_norm": 27.199996948242188,
|
| 75 |
+
"learning_rate": 1.0874316939890712e-05,
|
| 76 |
+
"loss": 4.0963,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.2304526748971194,
|
| 81 |
+
"grad_norm": 24.940441131591797,
|
| 82 |
+
"learning_rate": 1.2240437158469946e-05,
|
| 83 |
+
"loss": 4.1392,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.3676268861454046,
|
| 88 |
+
"grad_norm": 107.5506820678711,
|
| 89 |
+
"learning_rate": 1.3606557377049181e-05,
|
| 90 |
+
"loss": 3.897,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.50480109739369,
|
| 95 |
+
"grad_norm": 57.09001922607422,
|
| 96 |
+
"learning_rate": 1.4972677595628417e-05,
|
| 97 |
+
"loss": 3.6484,
|
| 98 |
+
"step": 275
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.6419753086419753,
|
| 102 |
+
"grad_norm": 41.39326858520508,
|
| 103 |
+
"learning_rate": 1.633879781420765e-05,
|
| 104 |
+
"loss": 3.1905,
|
| 105 |
+
"step": 300
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.7791495198902605,
|
| 109 |
+
"grad_norm": 55.287750244140625,
|
| 110 |
+
"learning_rate": 1.7704918032786887e-05,
|
| 111 |
+
"loss": 3.0062,
|
| 112 |
+
"step": 325
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 1.916323731138546,
|
| 116 |
+
"grad_norm": 46.43519973754883,
|
| 117 |
+
"learning_rate": 1.9071038251366122e-05,
|
| 118 |
+
"loss": 2.6752,
|
| 119 |
+
"step": 350
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 2.0,
|
| 123 |
+
"eval_accuracy": 0.7886855241264559,
|
| 124 |
+
"eval_loss": 0.5525600910186768,
|
| 125 |
+
"eval_macro_f1": 0.7845924270049904,
|
| 126 |
+
"eval_runtime": 1.2021,
|
| 127 |
+
"eval_samples_per_second": 499.972,
|
| 128 |
+
"eval_steps_per_second": 31.612,
|
| 129 |
+
"eval_weighted_f1": 0.785228326850597,
|
| 130 |
+
"step": 366
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.049382716049383,
|
| 134 |
+
"grad_norm": 26.568479537963867,
|
| 135 |
+
"learning_rate": 1.995142683667274e-05,
|
| 136 |
+
"loss": 2.4311,
|
| 137 |
+
"step": 375
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 2.186556927297668,
|
| 141 |
+
"grad_norm": 56.59968185424805,
|
| 142 |
+
"learning_rate": 1.9799635701275045e-05,
|
| 143 |
+
"loss": 2.1499,
|
| 144 |
+
"step": 400
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.3237311385459534,
|
| 148 |
+
"grad_norm": 148.46084594726562,
|
| 149 |
+
"learning_rate": 1.9647844565877354e-05,
|
| 150 |
+
"loss": 2.1125,
|
| 151 |
+
"step": 425
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.460905349794239,
|
| 155 |
+
"grad_norm": 234.44358825683594,
|
| 156 |
+
"learning_rate": 1.9496053430479663e-05,
|
| 157 |
+
"loss": 2.1165,
|
| 158 |
+
"step": 450
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.598079561042524,
|
| 162 |
+
"grad_norm": 73.21241760253906,
|
| 163 |
+
"learning_rate": 1.934426229508197e-05,
|
| 164 |
+
"loss": 1.973,
|
| 165 |
+
"step": 475
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.7352537722908092,
|
| 169 |
+
"grad_norm": 32.86597442626953,
|
| 170 |
+
"learning_rate": 1.9192471159684274e-05,
|
| 171 |
+
"loss": 2.1793,
|
| 172 |
+
"step": 500
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 2.8724279835390947,
|
| 176 |
+
"grad_norm": 45.75342559814453,
|
| 177 |
+
"learning_rate": 1.9040680024286583e-05,
|
| 178 |
+
"loss": 1.9802,
|
| 179 |
+
"step": 525
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 3.0,
|
| 183 |
+
"eval_accuracy": 0.8302828618968386,
|
| 184 |
+
"eval_loss": 0.4791373610496521,
|
| 185 |
+
"eval_macro_f1": 0.8272674264938796,
|
| 186 |
+
"eval_runtime": 1.2439,
|
| 187 |
+
"eval_samples_per_second": 483.145,
|
| 188 |
+
"eval_steps_per_second": 30.548,
|
| 189 |
+
"eval_weighted_f1": 0.8279277616442157,
|
| 190 |
+
"step": 549
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 3.0054869684499312,
|
| 194 |
+
"grad_norm": 47.666439056396484,
|
| 195 |
+
"learning_rate": 1.888888888888889e-05,
|
| 196 |
+
"loss": 1.8481,
|
| 197 |
+
"step": 550
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 3.1426611796982167,
|
| 201 |
+
"grad_norm": 63.70184326171875,
|
| 202 |
+
"learning_rate": 1.8737097753491198e-05,
|
| 203 |
+
"loss": 1.4317,
|
| 204 |
+
"step": 575
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.279835390946502,
|
| 208 |
+
"grad_norm": 42.282135009765625,
|
| 209 |
+
"learning_rate": 1.8585306618093504e-05,
|
| 210 |
+
"loss": 1.3833,
|
| 211 |
+
"step": 600
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.4170096021947876,
|
| 215 |
+
"grad_norm": 50.52495193481445,
|
| 216 |
+
"learning_rate": 1.8433515482695813e-05,
|
| 217 |
+
"loss": 1.2621,
|
| 218 |
+
"step": 625
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.5541838134430725,
|
| 222 |
+
"grad_norm": 52.09380340576172,
|
| 223 |
+
"learning_rate": 1.828172434729812e-05,
|
| 224 |
+
"loss": 1.5077,
|
| 225 |
+
"step": 650
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.691358024691358,
|
| 229 |
+
"grad_norm": 61.65653991699219,
|
| 230 |
+
"learning_rate": 1.8129933211900424e-05,
|
| 231 |
+
"loss": 1.4412,
|
| 232 |
+
"step": 675
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.8285322359396434,
|
| 236 |
+
"grad_norm": 41.13845443725586,
|
| 237 |
+
"learning_rate": 1.7978142076502733e-05,
|
| 238 |
+
"loss": 1.459,
|
| 239 |
+
"step": 700
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 3.9657064471879284,
|
| 243 |
+
"grad_norm": 33.82706069946289,
|
| 244 |
+
"learning_rate": 1.7826350941105042e-05,
|
| 245 |
+
"loss": 1.4888,
|
| 246 |
+
"step": 725
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 4.0,
|
| 250 |
+
"eval_accuracy": 0.8053244592346089,
|
| 251 |
+
"eval_loss": 0.5439836978912354,
|
| 252 |
+
"eval_macro_f1": 0.805808150353419,
|
| 253 |
+
"eval_runtime": 1.264,
|
| 254 |
+
"eval_samples_per_second": 475.463,
|
| 255 |
+
"eval_steps_per_second": 30.063,
|
| 256 |
+
"eval_weighted_f1": 0.8063022032212204,
|
| 257 |
+
"step": 732
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 4.098765432098766,
|
| 261 |
+
"grad_norm": 35.514225006103516,
|
| 262 |
+
"learning_rate": 1.7674559805707348e-05,
|
| 263 |
+
"loss": 1.3508,
|
| 264 |
+
"step": 750
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.23593964334705,
|
| 268 |
+
"grad_norm": 69.92938995361328,
|
| 269 |
+
"learning_rate": 1.7522768670309657e-05,
|
| 270 |
+
"loss": 0.9213,
|
| 271 |
+
"step": 775
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 4.373113854595336,
|
| 275 |
+
"grad_norm": 42.172706604003906,
|
| 276 |
+
"learning_rate": 1.7370977534911962e-05,
|
| 277 |
+
"loss": 0.9194,
|
| 278 |
+
"step": 800
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 4.510288065843621,
|
| 282 |
+
"grad_norm": 75.66331481933594,
|
| 283 |
+
"learning_rate": 1.7219186399514268e-05,
|
| 284 |
+
"loss": 1.0388,
|
| 285 |
+
"step": 825
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 4.647462277091907,
|
| 289 |
+
"grad_norm": 165.6358642578125,
|
| 290 |
+
"learning_rate": 1.7067395264116577e-05,
|
| 291 |
+
"loss": 1.1328,
|
| 292 |
+
"step": 850
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 4.784636488340192,
|
| 296 |
+
"grad_norm": 24.87302017211914,
|
| 297 |
+
"learning_rate": 1.6915604128718883e-05,
|
| 298 |
+
"loss": 0.9332,
|
| 299 |
+
"step": 875
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 4.921810699588478,
|
| 303 |
+
"grad_norm": 41.34330368041992,
|
| 304 |
+
"learning_rate": 1.6763812993321192e-05,
|
| 305 |
+
"loss": 1.0243,
|
| 306 |
+
"step": 900
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 5.0,
|
| 310 |
+
"eval_accuracy": 0.8519134775374376,
|
| 311 |
+
"eval_loss": 0.5014016032218933,
|
| 312 |
+
"eval_macro_f1": 0.850143080906351,
|
| 313 |
+
"eval_runtime": 1.3505,
|
| 314 |
+
"eval_samples_per_second": 445.027,
|
| 315 |
+
"eval_steps_per_second": 28.138,
|
| 316 |
+
"eval_weighted_f1": 0.8504764894803195,
|
| 317 |
+
"step": 915
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 5.054869684499314,
|
| 321 |
+
"grad_norm": 158.2794952392578,
|
| 322 |
+
"learning_rate": 1.66120218579235e-05,
|
| 323 |
+
"loss": 0.8075,
|
| 324 |
+
"step": 925
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 5.1920438957476,
|
| 328 |
+
"grad_norm": 22.857133865356445,
|
| 329 |
+
"learning_rate": 1.6460230722525807e-05,
|
| 330 |
+
"loss": 0.6515,
|
| 331 |
+
"step": 950
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.329218106995885,
|
| 335 |
+
"grad_norm": 38.806854248046875,
|
| 336 |
+
"learning_rate": 1.6308439587128112e-05,
|
| 337 |
+
"loss": 0.5644,
|
| 338 |
+
"step": 975
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 5.4663923182441705,
|
| 342 |
+
"grad_norm": 92.0794906616211,
|
| 343 |
+
"learning_rate": 1.615664845173042e-05,
|
| 344 |
+
"loss": 0.6874,
|
| 345 |
+
"step": 1000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.603566529492455,
|
| 349 |
+
"grad_norm": 59.99885177612305,
|
| 350 |
+
"learning_rate": 1.6004857316332727e-05,
|
| 351 |
+
"loss": 0.6772,
|
| 352 |
+
"step": 1025
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 5.7407407407407405,
|
| 356 |
+
"grad_norm": 92.39310455322266,
|
| 357 |
+
"learning_rate": 1.5853066180935036e-05,
|
| 358 |
+
"loss": 0.8236,
|
| 359 |
+
"step": 1050
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 5.877914951989026,
|
| 363 |
+
"grad_norm": 51.172340393066406,
|
| 364 |
+
"learning_rate": 1.570127504553734e-05,
|
| 365 |
+
"loss": 0.6569,
|
| 366 |
+
"step": 1075
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 6.0,
|
| 370 |
+
"eval_accuracy": 0.8435940099833611,
|
| 371 |
+
"eval_loss": 0.5641887187957764,
|
| 372 |
+
"eval_macro_f1": 0.8435841638811917,
|
| 373 |
+
"eval_runtime": 1.1003,
|
| 374 |
+
"eval_samples_per_second": 546.235,
|
| 375 |
+
"eval_steps_per_second": 34.537,
|
| 376 |
+
"eval_weighted_f1": 0.8433240244011455,
|
| 377 |
+
"step": 1098
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 6.0109739368998625,
|
| 381 |
+
"grad_norm": 71.54668426513672,
|
| 382 |
+
"learning_rate": 1.5549483910139647e-05,
|
| 383 |
+
"loss": 0.9109,
|
| 384 |
+
"step": 1100
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 6.148148148148148,
|
| 388 |
+
"grad_norm": 57.96711730957031,
|
| 389 |
+
"learning_rate": 1.5397692774741956e-05,
|
| 390 |
+
"loss": 0.6454,
|
| 391 |
+
"step": 1125
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 6.285322359396433,
|
| 395 |
+
"grad_norm": 95.6554183959961,
|
| 396 |
+
"learning_rate": 1.5245901639344264e-05,
|
| 397 |
+
"loss": 0.5632,
|
| 398 |
+
"step": 1150
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 6.422496570644719,
|
| 402 |
+
"grad_norm": 123.9430923461914,
|
| 403 |
+
"learning_rate": 1.5094110503946571e-05,
|
| 404 |
+
"loss": 0.6786,
|
| 405 |
+
"step": 1175
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 6.559670781893004,
|
| 409 |
+
"grad_norm": 33.93818283081055,
|
| 410 |
+
"learning_rate": 1.4942319368548878e-05,
|
| 411 |
+
"loss": 0.3503,
|
| 412 |
+
"step": 1200
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 6.69684499314129,
|
| 416 |
+
"grad_norm": 25.84504508972168,
|
| 417 |
+
"learning_rate": 1.4790528233151184e-05,
|
| 418 |
+
"loss": 0.3905,
|
| 419 |
+
"step": 1225
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 6.834019204389575,
|
| 423 |
+
"grad_norm": 128.41444396972656,
|
| 424 |
+
"learning_rate": 1.4638737097753493e-05,
|
| 425 |
+
"loss": 0.4844,
|
| 426 |
+
"step": 1250
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 6.97119341563786,
|
| 430 |
+
"grad_norm": 64.23560333251953,
|
| 431 |
+
"learning_rate": 1.44869459623558e-05,
|
| 432 |
+
"loss": 0.5959,
|
| 433 |
+
"step": 1275
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 7.0,
|
| 437 |
+
"eval_accuracy": 0.8469217970049917,
|
| 438 |
+
"eval_loss": 0.6252807974815369,
|
| 439 |
+
"eval_macro_f1": 0.8465529712381098,
|
| 440 |
+
"eval_runtime": 1.0672,
|
| 441 |
+
"eval_samples_per_second": 563.172,
|
| 442 |
+
"eval_steps_per_second": 35.608,
|
| 443 |
+
"eval_weighted_f1": 0.846684634982057,
|
| 444 |
+
"step": 1281
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 7.104252400548697,
|
| 448 |
+
"grad_norm": 49.21171951293945,
|
| 449 |
+
"learning_rate": 1.4335154826958106e-05,
|
| 450 |
+
"loss": 0.2958,
|
| 451 |
+
"step": 1300
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 7.2414266117969825,
|
| 455 |
+
"grad_norm": 0.9688119888305664,
|
| 456 |
+
"learning_rate": 1.4183363691560415e-05,
|
| 457 |
+
"loss": 0.3732,
|
| 458 |
+
"step": 1325
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 7.378600823045267,
|
| 462 |
+
"grad_norm": 6.383370876312256,
|
| 463 |
+
"learning_rate": 1.403157255616272e-05,
|
| 464 |
+
"loss": 0.4186,
|
| 465 |
+
"step": 1350
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 7.5157750342935525,
|
| 469 |
+
"grad_norm": 73.21456146240234,
|
| 470 |
+
"learning_rate": 1.3879781420765028e-05,
|
| 471 |
+
"loss": 0.5495,
|
| 472 |
+
"step": 1375
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 7.652949245541838,
|
| 476 |
+
"grad_norm": 4.817716121673584,
|
| 477 |
+
"learning_rate": 1.3727990285367337e-05,
|
| 478 |
+
"loss": 0.5761,
|
| 479 |
+
"step": 1400
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 7.790123456790123,
|
| 483 |
+
"grad_norm": 85.5977554321289,
|
| 484 |
+
"learning_rate": 1.3576199149969643e-05,
|
| 485 |
+
"loss": 0.3991,
|
| 486 |
+
"step": 1425
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 7.927297668038409,
|
| 490 |
+
"grad_norm": 29.113855361938477,
|
| 491 |
+
"learning_rate": 1.342440801457195e-05,
|
| 492 |
+
"loss": 0.3994,
|
| 493 |
+
"step": 1450
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 8.0,
|
| 497 |
+
"eval_accuracy": 0.8386023294509152,
|
| 498 |
+
"eval_loss": 0.8299993276596069,
|
| 499 |
+
"eval_macro_f1": 0.8371540893701558,
|
| 500 |
+
"eval_runtime": 1.2734,
|
| 501 |
+
"eval_samples_per_second": 471.961,
|
| 502 |
+
"eval_steps_per_second": 29.841,
|
| 503 |
+
"eval_weighted_f1": 0.8373590114575822,
|
| 504 |
+
"step": 1464
|
| 505 |
+
}
|
| 506 |
+
],
|
| 507 |
+
"logging_steps": 25,
|
| 508 |
+
"max_steps": 3660,
|
| 509 |
+
"num_input_tokens_seen": 0,
|
| 510 |
+
"num_train_epochs": 20,
|
| 511 |
+
"save_steps": 500,
|
| 512 |
+
"stateful_callbacks": {
|
| 513 |
+
"EarlyStoppingCallback": {
|
| 514 |
+
"args": {
|
| 515 |
+
"early_stopping_patience": 3,
|
| 516 |
+
"early_stopping_threshold": 0.0
|
| 517 |
+
},
|
| 518 |
+
"attributes": {
|
| 519 |
+
"early_stopping_patience_counter": 3
|
| 520 |
+
}
|
| 521 |
+
},
|
| 522 |
+
"TrainerControl": {
|
| 523 |
+
"args": {
|
| 524 |
+
"should_epoch_stop": false,
|
| 525 |
+
"should_evaluate": false,
|
| 526 |
+
"should_log": false,
|
| 527 |
+
"should_save": true,
|
| 528 |
+
"should_training_stop": true
|
| 529 |
+
},
|
| 530 |
+
"attributes": {}
|
| 531 |
+
}
|
| 532 |
+
},
|
| 533 |
+
"total_flos": 0.0,
|
| 534 |
+
"train_batch_size": 8,
|
| 535 |
+
"trial_name": null,
|
| 536 |
+
"trial_params": null
|
| 537 |
+
}
|
ood-vifactcheck-bamibert-source-source-03-s43-20260829T010007Z-3858bb98/trainer/checkpoint-1647/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-source-source-03-s43-20260829T010007Z-3858bb98/trainer/checkpoint-1647/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-vifactcheck-bamibert-source-source-03-s43-20260829T010007Z-3858bb98/trainer/checkpoint-1647/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-source-source-03-s43-20260829T010007Z-3858bb98/trainer/checkpoint-1647/trainer_state.json
ADDED
|
@@ -0,0 +1,597 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1647,
|
| 3 |
+
"best_metric": 0.854686035805384,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-source-source-03-s43-20260829T010007Z-3858bb98/trainer/checkpoint-1647",
|
| 5 |
+
"epoch": 9.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1647,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.13717421124828533,
|
| 14 |
+
"grad_norm": 12.64020824432373,
|
| 15 |
+
"learning_rate": 1.3114754098360657e-06,
|
| 16 |
+
"loss": 4.4352,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.27434842249657065,
|
| 21 |
+
"grad_norm": 12.162772178649902,
|
| 22 |
+
"learning_rate": 2.677595628415301e-06,
|
| 23 |
+
"loss": 4.3927,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.411522633744856,
|
| 28 |
+
"grad_norm": 51.812076568603516,
|
| 29 |
+
"learning_rate": 4.043715846994536e-06,
|
| 30 |
+
"loss": 4.4052,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5486968449931413,
|
| 35 |
+
"grad_norm": 17.337196350097656,
|
| 36 |
+
"learning_rate": 5.409836065573772e-06,
|
| 37 |
+
"loss": 4.3477,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.6858710562414266,
|
| 42 |
+
"grad_norm": 13.106736183166504,
|
| 43 |
+
"learning_rate": 6.775956284153006e-06,
|
| 44 |
+
"loss": 4.3765,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.823045267489712,
|
| 49 |
+
"grad_norm": 11.13673210144043,
|
| 50 |
+
"learning_rate": 8.14207650273224e-06,
|
| 51 |
+
"loss": 4.3634,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.9602194787379973,
|
| 56 |
+
"grad_norm": 14.513246536254883,
|
| 57 |
+
"learning_rate": 9.508196721311476e-06,
|
| 58 |
+
"loss": 4.2311,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 1.0,
|
| 63 |
+
"eval_accuracy": 0.4292845257903494,
|
| 64 |
+
"eval_loss": 1.0447561740875244,
|
| 65 |
+
"eval_macro_f1": 0.37862402125096256,
|
| 66 |
+
"eval_runtime": 1.1803,
|
| 67 |
+
"eval_samples_per_second": 509.195,
|
| 68 |
+
"eval_steps_per_second": 32.195,
|
| 69 |
+
"eval_weighted_f1": 0.38439928377619687,
|
| 70 |
+
"step": 183
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.093278463648834,
|
| 74 |
+
"grad_norm": 20.107376098632812,
|
| 75 |
+
"learning_rate": 1.0874316939890712e-05,
|
| 76 |
+
"loss": 3.9886,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.2304526748971194,
|
| 81 |
+
"grad_norm": 43.85777282714844,
|
| 82 |
+
"learning_rate": 1.2240437158469946e-05,
|
| 83 |
+
"loss": 3.7347,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.3676268861454046,
|
| 88 |
+
"grad_norm": 28.803104400634766,
|
| 89 |
+
"learning_rate": 1.3606557377049181e-05,
|
| 90 |
+
"loss": 3.6423,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.50480109739369,
|
| 95 |
+
"grad_norm": 43.92806625366211,
|
| 96 |
+
"learning_rate": 1.4972677595628417e-05,
|
| 97 |
+
"loss": 3.3707,
|
| 98 |
+
"step": 275
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.6419753086419753,
|
| 102 |
+
"grad_norm": 39.72669982910156,
|
| 103 |
+
"learning_rate": 1.633879781420765e-05,
|
| 104 |
+
"loss": 3.1304,
|
| 105 |
+
"step": 300
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.7791495198902605,
|
| 109 |
+
"grad_norm": 34.46350860595703,
|
| 110 |
+
"learning_rate": 1.7704918032786887e-05,
|
| 111 |
+
"loss": 3.0941,
|
| 112 |
+
"step": 325
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 1.916323731138546,
|
| 116 |
+
"grad_norm": 30.932470321655273,
|
| 117 |
+
"learning_rate": 1.9071038251366122e-05,
|
| 118 |
+
"loss": 2.7224,
|
| 119 |
+
"step": 350
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 2.0,
|
| 123 |
+
"eval_accuracy": 0.7820299500831946,
|
| 124 |
+
"eval_loss": 0.5336191058158875,
|
| 125 |
+
"eval_macro_f1": 0.7821066434472536,
|
| 126 |
+
"eval_runtime": 1.1326,
|
| 127 |
+
"eval_samples_per_second": 530.629,
|
| 128 |
+
"eval_steps_per_second": 33.551,
|
| 129 |
+
"eval_weighted_f1": 0.7822600477679263,
|
| 130 |
+
"step": 366
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.049382716049383,
|
| 134 |
+
"grad_norm": 49.00942611694336,
|
| 135 |
+
"learning_rate": 1.995142683667274e-05,
|
| 136 |
+
"loss": 2.3899,
|
| 137 |
+
"step": 375
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 2.186556927297668,
|
| 141 |
+
"grad_norm": 76.6011734008789,
|
| 142 |
+
"learning_rate": 1.9799635701275045e-05,
|
| 143 |
+
"loss": 2.217,
|
| 144 |
+
"step": 400
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.3237311385459534,
|
| 148 |
+
"grad_norm": 44.3740119934082,
|
| 149 |
+
"learning_rate": 1.9647844565877354e-05,
|
| 150 |
+
"loss": 1.9672,
|
| 151 |
+
"step": 425
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.460905349794239,
|
| 155 |
+
"grad_norm": 49.174224853515625,
|
| 156 |
+
"learning_rate": 1.9496053430479663e-05,
|
| 157 |
+
"loss": 2.0068,
|
| 158 |
+
"step": 450
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.598079561042524,
|
| 162 |
+
"grad_norm": 50.935951232910156,
|
| 163 |
+
"learning_rate": 1.934426229508197e-05,
|
| 164 |
+
"loss": 2.1266,
|
| 165 |
+
"step": 475
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.7352537722908092,
|
| 169 |
+
"grad_norm": 53.36017990112305,
|
| 170 |
+
"learning_rate": 1.9192471159684274e-05,
|
| 171 |
+
"loss": 1.9879,
|
| 172 |
+
"step": 500
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 2.8724279835390947,
|
| 176 |
+
"grad_norm": 64.67961120605469,
|
| 177 |
+
"learning_rate": 1.9040680024286583e-05,
|
| 178 |
+
"loss": 1.981,
|
| 179 |
+
"step": 525
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 3.0,
|
| 183 |
+
"eval_accuracy": 0.826955074875208,
|
| 184 |
+
"eval_loss": 0.49085527658462524,
|
| 185 |
+
"eval_macro_f1": 0.8277884432712401,
|
| 186 |
+
"eval_runtime": 1.0665,
|
| 187 |
+
"eval_samples_per_second": 563.501,
|
| 188 |
+
"eval_steps_per_second": 35.629,
|
| 189 |
+
"eval_weighted_f1": 0.8280582297344963,
|
| 190 |
+
"step": 549
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 3.0054869684499312,
|
| 194 |
+
"grad_norm": 37.146427154541016,
|
| 195 |
+
"learning_rate": 1.888888888888889e-05,
|
| 196 |
+
"loss": 1.8236,
|
| 197 |
+
"step": 550
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 3.1426611796982167,
|
| 201 |
+
"grad_norm": 48.209388732910156,
|
| 202 |
+
"learning_rate": 1.8737097753491198e-05,
|
| 203 |
+
"loss": 1.3913,
|
| 204 |
+
"step": 575
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.279835390946502,
|
| 208 |
+
"grad_norm": 59.93416213989258,
|
| 209 |
+
"learning_rate": 1.8585306618093504e-05,
|
| 210 |
+
"loss": 1.4978,
|
| 211 |
+
"step": 600
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.4170096021947876,
|
| 215 |
+
"grad_norm": 70.67302703857422,
|
| 216 |
+
"learning_rate": 1.8433515482695813e-05,
|
| 217 |
+
"loss": 1.5313,
|
| 218 |
+
"step": 625
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.5541838134430725,
|
| 222 |
+
"grad_norm": 27.230087280273438,
|
| 223 |
+
"learning_rate": 1.828172434729812e-05,
|
| 224 |
+
"loss": 1.4778,
|
| 225 |
+
"step": 650
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.691358024691358,
|
| 229 |
+
"grad_norm": 21.121475219726562,
|
| 230 |
+
"learning_rate": 1.8129933211900424e-05,
|
| 231 |
+
"loss": 1.4431,
|
| 232 |
+
"step": 675
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.8285322359396434,
|
| 236 |
+
"grad_norm": 28.312164306640625,
|
| 237 |
+
"learning_rate": 1.7978142076502733e-05,
|
| 238 |
+
"loss": 1.331,
|
| 239 |
+
"step": 700
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 3.9657064471879284,
|
| 243 |
+
"grad_norm": 36.571372985839844,
|
| 244 |
+
"learning_rate": 1.7826350941105042e-05,
|
| 245 |
+
"loss": 1.2235,
|
| 246 |
+
"step": 725
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 4.0,
|
| 250 |
+
"eval_accuracy": 0.831946755407654,
|
| 251 |
+
"eval_loss": 0.48329150676727295,
|
| 252 |
+
"eval_macro_f1": 0.8313990475370167,
|
| 253 |
+
"eval_runtime": 1.0308,
|
| 254 |
+
"eval_samples_per_second": 583.053,
|
| 255 |
+
"eval_steps_per_second": 36.865,
|
| 256 |
+
"eval_weighted_f1": 0.8315986177012266,
|
| 257 |
+
"step": 732
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 4.098765432098766,
|
| 261 |
+
"grad_norm": 39.647552490234375,
|
| 262 |
+
"learning_rate": 1.7674559805707348e-05,
|
| 263 |
+
"loss": 1.0487,
|
| 264 |
+
"step": 750
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.23593964334705,
|
| 268 |
+
"grad_norm": 6.730307102203369,
|
| 269 |
+
"learning_rate": 1.7522768670309657e-05,
|
| 270 |
+
"loss": 1.0236,
|
| 271 |
+
"step": 775
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 4.373113854595336,
|
| 275 |
+
"grad_norm": 55.71085739135742,
|
| 276 |
+
"learning_rate": 1.7370977534911962e-05,
|
| 277 |
+
"loss": 0.8075,
|
| 278 |
+
"step": 800
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 4.510288065843621,
|
| 282 |
+
"grad_norm": 44.08854293823242,
|
| 283 |
+
"learning_rate": 1.7219186399514268e-05,
|
| 284 |
+
"loss": 1.0982,
|
| 285 |
+
"step": 825
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 4.647462277091907,
|
| 289 |
+
"grad_norm": 51.03850555419922,
|
| 290 |
+
"learning_rate": 1.7067395264116577e-05,
|
| 291 |
+
"loss": 0.8874,
|
| 292 |
+
"step": 850
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 4.784636488340192,
|
| 296 |
+
"grad_norm": 32.591983795166016,
|
| 297 |
+
"learning_rate": 1.6915604128718883e-05,
|
| 298 |
+
"loss": 0.9069,
|
| 299 |
+
"step": 875
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 4.921810699588478,
|
| 303 |
+
"grad_norm": 27.56911849975586,
|
| 304 |
+
"learning_rate": 1.6763812993321192e-05,
|
| 305 |
+
"loss": 1.0093,
|
| 306 |
+
"step": 900
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 5.0,
|
| 310 |
+
"eval_accuracy": 0.8336106489184693,
|
| 311 |
+
"eval_loss": 0.590367317199707,
|
| 312 |
+
"eval_macro_f1": 0.8339327049004468,
|
| 313 |
+
"eval_runtime": 1.0657,
|
| 314 |
+
"eval_samples_per_second": 563.946,
|
| 315 |
+
"eval_steps_per_second": 35.657,
|
| 316 |
+
"eval_weighted_f1": 0.834169586142105,
|
| 317 |
+
"step": 915
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 5.054869684499314,
|
| 321 |
+
"grad_norm": 30.687345504760742,
|
| 322 |
+
"learning_rate": 1.66120218579235e-05,
|
| 323 |
+
"loss": 0.837,
|
| 324 |
+
"step": 925
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 5.1920438957476,
|
| 328 |
+
"grad_norm": 67.07527923583984,
|
| 329 |
+
"learning_rate": 1.6460230722525807e-05,
|
| 330 |
+
"loss": 0.6347,
|
| 331 |
+
"step": 950
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.329218106995885,
|
| 335 |
+
"grad_norm": 54.86009979248047,
|
| 336 |
+
"learning_rate": 1.6308439587128112e-05,
|
| 337 |
+
"loss": 0.8212,
|
| 338 |
+
"step": 975
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 5.4663923182441705,
|
| 342 |
+
"grad_norm": 114.8167953491211,
|
| 343 |
+
"learning_rate": 1.615664845173042e-05,
|
| 344 |
+
"loss": 0.7812,
|
| 345 |
+
"step": 1000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.603566529492455,
|
| 349 |
+
"grad_norm": 58.114410400390625,
|
| 350 |
+
"learning_rate": 1.6004857316332727e-05,
|
| 351 |
+
"loss": 0.7166,
|
| 352 |
+
"step": 1025
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 5.7407407407407405,
|
| 356 |
+
"grad_norm": 105.2366943359375,
|
| 357 |
+
"learning_rate": 1.5853066180935036e-05,
|
| 358 |
+
"loss": 0.5473,
|
| 359 |
+
"step": 1050
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 5.877914951989026,
|
| 363 |
+
"grad_norm": 2.7807657718658447,
|
| 364 |
+
"learning_rate": 1.570127504553734e-05,
|
| 365 |
+
"loss": 0.588,
|
| 366 |
+
"step": 1075
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 6.0,
|
| 370 |
+
"eval_accuracy": 0.8202995008319468,
|
| 371 |
+
"eval_loss": 0.767414391040802,
|
| 372 |
+
"eval_macro_f1": 0.818518146732076,
|
| 373 |
+
"eval_runtime": 1.087,
|
| 374 |
+
"eval_samples_per_second": 552.922,
|
| 375 |
+
"eval_steps_per_second": 34.96,
|
| 376 |
+
"eval_weighted_f1": 0.8186086828490861,
|
| 377 |
+
"step": 1098
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 6.0109739368998625,
|
| 381 |
+
"grad_norm": 161.01165771484375,
|
| 382 |
+
"learning_rate": 1.5549483910139647e-05,
|
| 383 |
+
"loss": 0.6709,
|
| 384 |
+
"step": 1100
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 6.148148148148148,
|
| 388 |
+
"grad_norm": 76.24903106689453,
|
| 389 |
+
"learning_rate": 1.5397692774741956e-05,
|
| 390 |
+
"loss": 0.5097,
|
| 391 |
+
"step": 1125
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 6.285322359396433,
|
| 395 |
+
"grad_norm": 7.166459560394287,
|
| 396 |
+
"learning_rate": 1.5245901639344264e-05,
|
| 397 |
+
"loss": 0.5035,
|
| 398 |
+
"step": 1150
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 6.422496570644719,
|
| 402 |
+
"grad_norm": 1.4280349016189575,
|
| 403 |
+
"learning_rate": 1.5094110503946571e-05,
|
| 404 |
+
"loss": 0.6442,
|
| 405 |
+
"step": 1175
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 6.559670781893004,
|
| 409 |
+
"grad_norm": 34.56960678100586,
|
| 410 |
+
"learning_rate": 1.4942319368548878e-05,
|
| 411 |
+
"loss": 0.7182,
|
| 412 |
+
"step": 1200
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 6.69684499314129,
|
| 416 |
+
"grad_norm": 8.275811195373535,
|
| 417 |
+
"learning_rate": 1.4790528233151184e-05,
|
| 418 |
+
"loss": 0.4541,
|
| 419 |
+
"step": 1225
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 6.834019204389575,
|
| 423 |
+
"grad_norm": 87.94701385498047,
|
| 424 |
+
"learning_rate": 1.4638737097753493e-05,
|
| 425 |
+
"loss": 0.5964,
|
| 426 |
+
"step": 1250
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 6.97119341563786,
|
| 430 |
+
"grad_norm": 10.228333473205566,
|
| 431 |
+
"learning_rate": 1.44869459623558e-05,
|
| 432 |
+
"loss": 0.4256,
|
| 433 |
+
"step": 1275
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 7.0,
|
| 437 |
+
"eval_accuracy": 0.8369384359400999,
|
| 438 |
+
"eval_loss": 0.7751681804656982,
|
| 439 |
+
"eval_macro_f1": 0.8373943368549625,
|
| 440 |
+
"eval_runtime": 1.064,
|
| 441 |
+
"eval_samples_per_second": 564.833,
|
| 442 |
+
"eval_steps_per_second": 35.713,
|
| 443 |
+
"eval_weighted_f1": 0.8374212132134473,
|
| 444 |
+
"step": 1281
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 7.104252400548697,
|
| 448 |
+
"grad_norm": 31.649810791015625,
|
| 449 |
+
"learning_rate": 1.4335154826958106e-05,
|
| 450 |
+
"loss": 0.4033,
|
| 451 |
+
"step": 1300
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 7.2414266117969825,
|
| 455 |
+
"grad_norm": 7.597397804260254,
|
| 456 |
+
"learning_rate": 1.4183363691560415e-05,
|
| 457 |
+
"loss": 0.3152,
|
| 458 |
+
"step": 1325
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 7.378600823045267,
|
| 462 |
+
"grad_norm": 17.44788932800293,
|
| 463 |
+
"learning_rate": 1.403157255616272e-05,
|
| 464 |
+
"loss": 0.3159,
|
| 465 |
+
"step": 1350
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 7.5157750342935525,
|
| 469 |
+
"grad_norm": 3.1336617469787598,
|
| 470 |
+
"learning_rate": 1.3879781420765028e-05,
|
| 471 |
+
"loss": 0.4866,
|
| 472 |
+
"step": 1375
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 7.652949245541838,
|
| 476 |
+
"grad_norm": 6.395193099975586,
|
| 477 |
+
"learning_rate": 1.3727990285367337e-05,
|
| 478 |
+
"loss": 0.4649,
|
| 479 |
+
"step": 1400
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 7.790123456790123,
|
| 483 |
+
"grad_norm": 25.774215698242188,
|
| 484 |
+
"learning_rate": 1.3576199149969643e-05,
|
| 485 |
+
"loss": 0.4685,
|
| 486 |
+
"step": 1425
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 7.927297668038409,
|
| 490 |
+
"grad_norm": 98.32450103759766,
|
| 491 |
+
"learning_rate": 1.342440801457195e-05,
|
| 492 |
+
"loss": 0.5285,
|
| 493 |
+
"step": 1450
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 8.0,
|
| 497 |
+
"eval_accuracy": 0.8435940099833611,
|
| 498 |
+
"eval_loss": 0.8015182614326477,
|
| 499 |
+
"eval_macro_f1": 0.8442057072589121,
|
| 500 |
+
"eval_runtime": 1.236,
|
| 501 |
+
"eval_samples_per_second": 486.261,
|
| 502 |
+
"eval_steps_per_second": 30.745,
|
| 503 |
+
"eval_weighted_f1": 0.8440409596055877,
|
| 504 |
+
"step": 1464
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 8.060356652949245,
|
| 508 |
+
"grad_norm": 80.06298828125,
|
| 509 |
+
"learning_rate": 1.3272616879174259e-05,
|
| 510 |
+
"loss": 0.3845,
|
| 511 |
+
"step": 1475
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 8.197530864197532,
|
| 515 |
+
"grad_norm": 94.94721984863281,
|
| 516 |
+
"learning_rate": 1.3120825743776565e-05,
|
| 517 |
+
"loss": 0.2777,
|
| 518 |
+
"step": 1500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 8.334705075445816,
|
| 522 |
+
"grad_norm": 58.17174530029297,
|
| 523 |
+
"learning_rate": 1.2969034608378872e-05,
|
| 524 |
+
"loss": 0.4543,
|
| 525 |
+
"step": 1525
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 8.4718792866941,
|
| 529 |
+
"grad_norm": 113.3753433227539,
|
| 530 |
+
"learning_rate": 1.2817243472981178e-05,
|
| 531 |
+
"loss": 0.5612,
|
| 532 |
+
"step": 1550
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 8.609053497942387,
|
| 536 |
+
"grad_norm": 9.769021987915039,
|
| 537 |
+
"learning_rate": 1.2665452337583487e-05,
|
| 538 |
+
"loss": 0.27,
|
| 539 |
+
"step": 1575
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 8.746227709190672,
|
| 543 |
+
"grad_norm": 23.138566970825195,
|
| 544 |
+
"learning_rate": 1.2513661202185794e-05,
|
| 545 |
+
"loss": 0.3435,
|
| 546 |
+
"step": 1600
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 8.883401920438958,
|
| 550 |
+
"grad_norm": 1.9317066669464111,
|
| 551 |
+
"learning_rate": 1.23618700667881e-05,
|
| 552 |
+
"loss": 0.3948,
|
| 553 |
+
"step": 1625
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 9.0,
|
| 557 |
+
"eval_accuracy": 0.8552412645590682,
|
| 558 |
+
"eval_loss": 0.8615436553955078,
|
| 559 |
+
"eval_macro_f1": 0.854686035805384,
|
| 560 |
+
"eval_runtime": 1.0462,
|
| 561 |
+
"eval_samples_per_second": 574.434,
|
| 562 |
+
"eval_steps_per_second": 36.32,
|
| 563 |
+
"eval_weighted_f1": 0.8548245021423386,
|
| 564 |
+
"step": 1647
|
| 565 |
+
}
|
| 566 |
+
],
|
| 567 |
+
"logging_steps": 25,
|
| 568 |
+
"max_steps": 3660,
|
| 569 |
+
"num_input_tokens_seen": 0,
|
| 570 |
+
"num_train_epochs": 20,
|
| 571 |
+
"save_steps": 500,
|
| 572 |
+
"stateful_callbacks": {
|
| 573 |
+
"EarlyStoppingCallback": {
|
| 574 |
+
"args": {
|
| 575 |
+
"early_stopping_patience": 3,
|
| 576 |
+
"early_stopping_threshold": 0.0
|
| 577 |
+
},
|
| 578 |
+
"attributes": {
|
| 579 |
+
"early_stopping_patience_counter": 0
|
| 580 |
+
}
|
| 581 |
+
},
|
| 582 |
+
"TrainerControl": {
|
| 583 |
+
"args": {
|
| 584 |
+
"should_epoch_stop": false,
|
| 585 |
+
"should_evaluate": false,
|
| 586 |
+
"should_log": false,
|
| 587 |
+
"should_save": true,
|
| 588 |
+
"should_training_stop": false
|
| 589 |
+
},
|
| 590 |
+
"attributes": {}
|
| 591 |
+
}
|
| 592 |
+
},
|
| 593 |
+
"total_flos": 0.0,
|
| 594 |
+
"train_batch_size": 8,
|
| 595 |
+
"trial_name": null,
|
| 596 |
+
"trial_params": null
|
| 597 |
+
}
|
ood-vifactcheck-bamibert-source-source-03-s43-20260829T010007Z-3858bb98/trainer/checkpoint-2196/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-source-source-03-s43-20260829T010007Z-3858bb98/trainer/checkpoint-2196/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-vifactcheck-bamibert-source-source-03-s43-20260829T010007Z-3858bb98/trainer/checkpoint-2196/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-source-source-03-s43-20260829T010007Z-3858bb98/trainer/checkpoint-2196/trainer_state.json
ADDED
|
@@ -0,0 +1,784 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1647,
|
| 3 |
+
"best_metric": 0.854686035805384,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-source-source-03-s43-20260829T010007Z-3858bb98/trainer/checkpoint-1647",
|
| 5 |
+
"epoch": 12.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2196,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.13717421124828533,
|
| 14 |
+
"grad_norm": 12.64020824432373,
|
| 15 |
+
"learning_rate": 1.3114754098360657e-06,
|
| 16 |
+
"loss": 4.4352,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.27434842249657065,
|
| 21 |
+
"grad_norm": 12.162772178649902,
|
| 22 |
+
"learning_rate": 2.677595628415301e-06,
|
| 23 |
+
"loss": 4.3927,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.411522633744856,
|
| 28 |
+
"grad_norm": 51.812076568603516,
|
| 29 |
+
"learning_rate": 4.043715846994536e-06,
|
| 30 |
+
"loss": 4.4052,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5486968449931413,
|
| 35 |
+
"grad_norm": 17.337196350097656,
|
| 36 |
+
"learning_rate": 5.409836065573772e-06,
|
| 37 |
+
"loss": 4.3477,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.6858710562414266,
|
| 42 |
+
"grad_norm": 13.106736183166504,
|
| 43 |
+
"learning_rate": 6.775956284153006e-06,
|
| 44 |
+
"loss": 4.3765,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.823045267489712,
|
| 49 |
+
"grad_norm": 11.13673210144043,
|
| 50 |
+
"learning_rate": 8.14207650273224e-06,
|
| 51 |
+
"loss": 4.3634,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.9602194787379973,
|
| 56 |
+
"grad_norm": 14.513246536254883,
|
| 57 |
+
"learning_rate": 9.508196721311476e-06,
|
| 58 |
+
"loss": 4.2311,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 1.0,
|
| 63 |
+
"eval_accuracy": 0.4292845257903494,
|
| 64 |
+
"eval_loss": 1.0447561740875244,
|
| 65 |
+
"eval_macro_f1": 0.37862402125096256,
|
| 66 |
+
"eval_runtime": 1.1803,
|
| 67 |
+
"eval_samples_per_second": 509.195,
|
| 68 |
+
"eval_steps_per_second": 32.195,
|
| 69 |
+
"eval_weighted_f1": 0.38439928377619687,
|
| 70 |
+
"step": 183
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.093278463648834,
|
| 74 |
+
"grad_norm": 20.107376098632812,
|
| 75 |
+
"learning_rate": 1.0874316939890712e-05,
|
| 76 |
+
"loss": 3.9886,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.2304526748971194,
|
| 81 |
+
"grad_norm": 43.85777282714844,
|
| 82 |
+
"learning_rate": 1.2240437158469946e-05,
|
| 83 |
+
"loss": 3.7347,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.3676268861454046,
|
| 88 |
+
"grad_norm": 28.803104400634766,
|
| 89 |
+
"learning_rate": 1.3606557377049181e-05,
|
| 90 |
+
"loss": 3.6423,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.50480109739369,
|
| 95 |
+
"grad_norm": 43.92806625366211,
|
| 96 |
+
"learning_rate": 1.4972677595628417e-05,
|
| 97 |
+
"loss": 3.3707,
|
| 98 |
+
"step": 275
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.6419753086419753,
|
| 102 |
+
"grad_norm": 39.72669982910156,
|
| 103 |
+
"learning_rate": 1.633879781420765e-05,
|
| 104 |
+
"loss": 3.1304,
|
| 105 |
+
"step": 300
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.7791495198902605,
|
| 109 |
+
"grad_norm": 34.46350860595703,
|
| 110 |
+
"learning_rate": 1.7704918032786887e-05,
|
| 111 |
+
"loss": 3.0941,
|
| 112 |
+
"step": 325
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 1.916323731138546,
|
| 116 |
+
"grad_norm": 30.932470321655273,
|
| 117 |
+
"learning_rate": 1.9071038251366122e-05,
|
| 118 |
+
"loss": 2.7224,
|
| 119 |
+
"step": 350
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 2.0,
|
| 123 |
+
"eval_accuracy": 0.7820299500831946,
|
| 124 |
+
"eval_loss": 0.5336191058158875,
|
| 125 |
+
"eval_macro_f1": 0.7821066434472536,
|
| 126 |
+
"eval_runtime": 1.1326,
|
| 127 |
+
"eval_samples_per_second": 530.629,
|
| 128 |
+
"eval_steps_per_second": 33.551,
|
| 129 |
+
"eval_weighted_f1": 0.7822600477679263,
|
| 130 |
+
"step": 366
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.049382716049383,
|
| 134 |
+
"grad_norm": 49.00942611694336,
|
| 135 |
+
"learning_rate": 1.995142683667274e-05,
|
| 136 |
+
"loss": 2.3899,
|
| 137 |
+
"step": 375
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 2.186556927297668,
|
| 141 |
+
"grad_norm": 76.6011734008789,
|
| 142 |
+
"learning_rate": 1.9799635701275045e-05,
|
| 143 |
+
"loss": 2.217,
|
| 144 |
+
"step": 400
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.3237311385459534,
|
| 148 |
+
"grad_norm": 44.3740119934082,
|
| 149 |
+
"learning_rate": 1.9647844565877354e-05,
|
| 150 |
+
"loss": 1.9672,
|
| 151 |
+
"step": 425
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.460905349794239,
|
| 155 |
+
"grad_norm": 49.174224853515625,
|
| 156 |
+
"learning_rate": 1.9496053430479663e-05,
|
| 157 |
+
"loss": 2.0068,
|
| 158 |
+
"step": 450
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.598079561042524,
|
| 162 |
+
"grad_norm": 50.935951232910156,
|
| 163 |
+
"learning_rate": 1.934426229508197e-05,
|
| 164 |
+
"loss": 2.1266,
|
| 165 |
+
"step": 475
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.7352537722908092,
|
| 169 |
+
"grad_norm": 53.36017990112305,
|
| 170 |
+
"learning_rate": 1.9192471159684274e-05,
|
| 171 |
+
"loss": 1.9879,
|
| 172 |
+
"step": 500
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 2.8724279835390947,
|
| 176 |
+
"grad_norm": 64.67961120605469,
|
| 177 |
+
"learning_rate": 1.9040680024286583e-05,
|
| 178 |
+
"loss": 1.981,
|
| 179 |
+
"step": 525
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 3.0,
|
| 183 |
+
"eval_accuracy": 0.826955074875208,
|
| 184 |
+
"eval_loss": 0.49085527658462524,
|
| 185 |
+
"eval_macro_f1": 0.8277884432712401,
|
| 186 |
+
"eval_runtime": 1.0665,
|
| 187 |
+
"eval_samples_per_second": 563.501,
|
| 188 |
+
"eval_steps_per_second": 35.629,
|
| 189 |
+
"eval_weighted_f1": 0.8280582297344963,
|
| 190 |
+
"step": 549
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 3.0054869684499312,
|
| 194 |
+
"grad_norm": 37.146427154541016,
|
| 195 |
+
"learning_rate": 1.888888888888889e-05,
|
| 196 |
+
"loss": 1.8236,
|
| 197 |
+
"step": 550
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 3.1426611796982167,
|
| 201 |
+
"grad_norm": 48.209388732910156,
|
| 202 |
+
"learning_rate": 1.8737097753491198e-05,
|
| 203 |
+
"loss": 1.3913,
|
| 204 |
+
"step": 575
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.279835390946502,
|
| 208 |
+
"grad_norm": 59.93416213989258,
|
| 209 |
+
"learning_rate": 1.8585306618093504e-05,
|
| 210 |
+
"loss": 1.4978,
|
| 211 |
+
"step": 600
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.4170096021947876,
|
| 215 |
+
"grad_norm": 70.67302703857422,
|
| 216 |
+
"learning_rate": 1.8433515482695813e-05,
|
| 217 |
+
"loss": 1.5313,
|
| 218 |
+
"step": 625
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.5541838134430725,
|
| 222 |
+
"grad_norm": 27.230087280273438,
|
| 223 |
+
"learning_rate": 1.828172434729812e-05,
|
| 224 |
+
"loss": 1.4778,
|
| 225 |
+
"step": 650
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.691358024691358,
|
| 229 |
+
"grad_norm": 21.121475219726562,
|
| 230 |
+
"learning_rate": 1.8129933211900424e-05,
|
| 231 |
+
"loss": 1.4431,
|
| 232 |
+
"step": 675
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.8285322359396434,
|
| 236 |
+
"grad_norm": 28.312164306640625,
|
| 237 |
+
"learning_rate": 1.7978142076502733e-05,
|
| 238 |
+
"loss": 1.331,
|
| 239 |
+
"step": 700
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 3.9657064471879284,
|
| 243 |
+
"grad_norm": 36.571372985839844,
|
| 244 |
+
"learning_rate": 1.7826350941105042e-05,
|
| 245 |
+
"loss": 1.2235,
|
| 246 |
+
"step": 725
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 4.0,
|
| 250 |
+
"eval_accuracy": 0.831946755407654,
|
| 251 |
+
"eval_loss": 0.48329150676727295,
|
| 252 |
+
"eval_macro_f1": 0.8313990475370167,
|
| 253 |
+
"eval_runtime": 1.0308,
|
| 254 |
+
"eval_samples_per_second": 583.053,
|
| 255 |
+
"eval_steps_per_second": 36.865,
|
| 256 |
+
"eval_weighted_f1": 0.8315986177012266,
|
| 257 |
+
"step": 732
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 4.098765432098766,
|
| 261 |
+
"grad_norm": 39.647552490234375,
|
| 262 |
+
"learning_rate": 1.7674559805707348e-05,
|
| 263 |
+
"loss": 1.0487,
|
| 264 |
+
"step": 750
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.23593964334705,
|
| 268 |
+
"grad_norm": 6.730307102203369,
|
| 269 |
+
"learning_rate": 1.7522768670309657e-05,
|
| 270 |
+
"loss": 1.0236,
|
| 271 |
+
"step": 775
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 4.373113854595336,
|
| 275 |
+
"grad_norm": 55.71085739135742,
|
| 276 |
+
"learning_rate": 1.7370977534911962e-05,
|
| 277 |
+
"loss": 0.8075,
|
| 278 |
+
"step": 800
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 4.510288065843621,
|
| 282 |
+
"grad_norm": 44.08854293823242,
|
| 283 |
+
"learning_rate": 1.7219186399514268e-05,
|
| 284 |
+
"loss": 1.0982,
|
| 285 |
+
"step": 825
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 4.647462277091907,
|
| 289 |
+
"grad_norm": 51.03850555419922,
|
| 290 |
+
"learning_rate": 1.7067395264116577e-05,
|
| 291 |
+
"loss": 0.8874,
|
| 292 |
+
"step": 850
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 4.784636488340192,
|
| 296 |
+
"grad_norm": 32.591983795166016,
|
| 297 |
+
"learning_rate": 1.6915604128718883e-05,
|
| 298 |
+
"loss": 0.9069,
|
| 299 |
+
"step": 875
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 4.921810699588478,
|
| 303 |
+
"grad_norm": 27.56911849975586,
|
| 304 |
+
"learning_rate": 1.6763812993321192e-05,
|
| 305 |
+
"loss": 1.0093,
|
| 306 |
+
"step": 900
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 5.0,
|
| 310 |
+
"eval_accuracy": 0.8336106489184693,
|
| 311 |
+
"eval_loss": 0.590367317199707,
|
| 312 |
+
"eval_macro_f1": 0.8339327049004468,
|
| 313 |
+
"eval_runtime": 1.0657,
|
| 314 |
+
"eval_samples_per_second": 563.946,
|
| 315 |
+
"eval_steps_per_second": 35.657,
|
| 316 |
+
"eval_weighted_f1": 0.834169586142105,
|
| 317 |
+
"step": 915
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 5.054869684499314,
|
| 321 |
+
"grad_norm": 30.687345504760742,
|
| 322 |
+
"learning_rate": 1.66120218579235e-05,
|
| 323 |
+
"loss": 0.837,
|
| 324 |
+
"step": 925
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 5.1920438957476,
|
| 328 |
+
"grad_norm": 67.07527923583984,
|
| 329 |
+
"learning_rate": 1.6460230722525807e-05,
|
| 330 |
+
"loss": 0.6347,
|
| 331 |
+
"step": 950
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.329218106995885,
|
| 335 |
+
"grad_norm": 54.86009979248047,
|
| 336 |
+
"learning_rate": 1.6308439587128112e-05,
|
| 337 |
+
"loss": 0.8212,
|
| 338 |
+
"step": 975
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 5.4663923182441705,
|
| 342 |
+
"grad_norm": 114.8167953491211,
|
| 343 |
+
"learning_rate": 1.615664845173042e-05,
|
| 344 |
+
"loss": 0.7812,
|
| 345 |
+
"step": 1000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.603566529492455,
|
| 349 |
+
"grad_norm": 58.114410400390625,
|
| 350 |
+
"learning_rate": 1.6004857316332727e-05,
|
| 351 |
+
"loss": 0.7166,
|
| 352 |
+
"step": 1025
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 5.7407407407407405,
|
| 356 |
+
"grad_norm": 105.2366943359375,
|
| 357 |
+
"learning_rate": 1.5853066180935036e-05,
|
| 358 |
+
"loss": 0.5473,
|
| 359 |
+
"step": 1050
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 5.877914951989026,
|
| 363 |
+
"grad_norm": 2.7807657718658447,
|
| 364 |
+
"learning_rate": 1.570127504553734e-05,
|
| 365 |
+
"loss": 0.588,
|
| 366 |
+
"step": 1075
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 6.0,
|
| 370 |
+
"eval_accuracy": 0.8202995008319468,
|
| 371 |
+
"eval_loss": 0.767414391040802,
|
| 372 |
+
"eval_macro_f1": 0.818518146732076,
|
| 373 |
+
"eval_runtime": 1.087,
|
| 374 |
+
"eval_samples_per_second": 552.922,
|
| 375 |
+
"eval_steps_per_second": 34.96,
|
| 376 |
+
"eval_weighted_f1": 0.8186086828490861,
|
| 377 |
+
"step": 1098
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 6.0109739368998625,
|
| 381 |
+
"grad_norm": 161.01165771484375,
|
| 382 |
+
"learning_rate": 1.5549483910139647e-05,
|
| 383 |
+
"loss": 0.6709,
|
| 384 |
+
"step": 1100
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 6.148148148148148,
|
| 388 |
+
"grad_norm": 76.24903106689453,
|
| 389 |
+
"learning_rate": 1.5397692774741956e-05,
|
| 390 |
+
"loss": 0.5097,
|
| 391 |
+
"step": 1125
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 6.285322359396433,
|
| 395 |
+
"grad_norm": 7.166459560394287,
|
| 396 |
+
"learning_rate": 1.5245901639344264e-05,
|
| 397 |
+
"loss": 0.5035,
|
| 398 |
+
"step": 1150
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 6.422496570644719,
|
| 402 |
+
"grad_norm": 1.4280349016189575,
|
| 403 |
+
"learning_rate": 1.5094110503946571e-05,
|
| 404 |
+
"loss": 0.6442,
|
| 405 |
+
"step": 1175
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 6.559670781893004,
|
| 409 |
+
"grad_norm": 34.56960678100586,
|
| 410 |
+
"learning_rate": 1.4942319368548878e-05,
|
| 411 |
+
"loss": 0.7182,
|
| 412 |
+
"step": 1200
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 6.69684499314129,
|
| 416 |
+
"grad_norm": 8.275811195373535,
|
| 417 |
+
"learning_rate": 1.4790528233151184e-05,
|
| 418 |
+
"loss": 0.4541,
|
| 419 |
+
"step": 1225
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 6.834019204389575,
|
| 423 |
+
"grad_norm": 87.94701385498047,
|
| 424 |
+
"learning_rate": 1.4638737097753493e-05,
|
| 425 |
+
"loss": 0.5964,
|
| 426 |
+
"step": 1250
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 6.97119341563786,
|
| 430 |
+
"grad_norm": 10.228333473205566,
|
| 431 |
+
"learning_rate": 1.44869459623558e-05,
|
| 432 |
+
"loss": 0.4256,
|
| 433 |
+
"step": 1275
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 7.0,
|
| 437 |
+
"eval_accuracy": 0.8369384359400999,
|
| 438 |
+
"eval_loss": 0.7751681804656982,
|
| 439 |
+
"eval_macro_f1": 0.8373943368549625,
|
| 440 |
+
"eval_runtime": 1.064,
|
| 441 |
+
"eval_samples_per_second": 564.833,
|
| 442 |
+
"eval_steps_per_second": 35.713,
|
| 443 |
+
"eval_weighted_f1": 0.8374212132134473,
|
| 444 |
+
"step": 1281
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 7.104252400548697,
|
| 448 |
+
"grad_norm": 31.649810791015625,
|
| 449 |
+
"learning_rate": 1.4335154826958106e-05,
|
| 450 |
+
"loss": 0.4033,
|
| 451 |
+
"step": 1300
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 7.2414266117969825,
|
| 455 |
+
"grad_norm": 7.597397804260254,
|
| 456 |
+
"learning_rate": 1.4183363691560415e-05,
|
| 457 |
+
"loss": 0.3152,
|
| 458 |
+
"step": 1325
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 7.378600823045267,
|
| 462 |
+
"grad_norm": 17.44788932800293,
|
| 463 |
+
"learning_rate": 1.403157255616272e-05,
|
| 464 |
+
"loss": 0.3159,
|
| 465 |
+
"step": 1350
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 7.5157750342935525,
|
| 469 |
+
"grad_norm": 3.1336617469787598,
|
| 470 |
+
"learning_rate": 1.3879781420765028e-05,
|
| 471 |
+
"loss": 0.4866,
|
| 472 |
+
"step": 1375
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 7.652949245541838,
|
| 476 |
+
"grad_norm": 6.395193099975586,
|
| 477 |
+
"learning_rate": 1.3727990285367337e-05,
|
| 478 |
+
"loss": 0.4649,
|
| 479 |
+
"step": 1400
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 7.790123456790123,
|
| 483 |
+
"grad_norm": 25.774215698242188,
|
| 484 |
+
"learning_rate": 1.3576199149969643e-05,
|
| 485 |
+
"loss": 0.4685,
|
| 486 |
+
"step": 1425
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 7.927297668038409,
|
| 490 |
+
"grad_norm": 98.32450103759766,
|
| 491 |
+
"learning_rate": 1.342440801457195e-05,
|
| 492 |
+
"loss": 0.5285,
|
| 493 |
+
"step": 1450
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 8.0,
|
| 497 |
+
"eval_accuracy": 0.8435940099833611,
|
| 498 |
+
"eval_loss": 0.8015182614326477,
|
| 499 |
+
"eval_macro_f1": 0.8442057072589121,
|
| 500 |
+
"eval_runtime": 1.236,
|
| 501 |
+
"eval_samples_per_second": 486.261,
|
| 502 |
+
"eval_steps_per_second": 30.745,
|
| 503 |
+
"eval_weighted_f1": 0.8440409596055877,
|
| 504 |
+
"step": 1464
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 8.060356652949245,
|
| 508 |
+
"grad_norm": 80.06298828125,
|
| 509 |
+
"learning_rate": 1.3272616879174259e-05,
|
| 510 |
+
"loss": 0.3845,
|
| 511 |
+
"step": 1475
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 8.197530864197532,
|
| 515 |
+
"grad_norm": 94.94721984863281,
|
| 516 |
+
"learning_rate": 1.3120825743776565e-05,
|
| 517 |
+
"loss": 0.2777,
|
| 518 |
+
"step": 1500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 8.334705075445816,
|
| 522 |
+
"grad_norm": 58.17174530029297,
|
| 523 |
+
"learning_rate": 1.2969034608378872e-05,
|
| 524 |
+
"loss": 0.4543,
|
| 525 |
+
"step": 1525
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 8.4718792866941,
|
| 529 |
+
"grad_norm": 113.3753433227539,
|
| 530 |
+
"learning_rate": 1.2817243472981178e-05,
|
| 531 |
+
"loss": 0.5612,
|
| 532 |
+
"step": 1550
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 8.609053497942387,
|
| 536 |
+
"grad_norm": 9.769021987915039,
|
| 537 |
+
"learning_rate": 1.2665452337583487e-05,
|
| 538 |
+
"loss": 0.27,
|
| 539 |
+
"step": 1575
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 8.746227709190672,
|
| 543 |
+
"grad_norm": 23.138566970825195,
|
| 544 |
+
"learning_rate": 1.2513661202185794e-05,
|
| 545 |
+
"loss": 0.3435,
|
| 546 |
+
"step": 1600
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 8.883401920438958,
|
| 550 |
+
"grad_norm": 1.9317066669464111,
|
| 551 |
+
"learning_rate": 1.23618700667881e-05,
|
| 552 |
+
"loss": 0.3948,
|
| 553 |
+
"step": 1625
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 9.0,
|
| 557 |
+
"eval_accuracy": 0.8552412645590682,
|
| 558 |
+
"eval_loss": 0.8615436553955078,
|
| 559 |
+
"eval_macro_f1": 0.854686035805384,
|
| 560 |
+
"eval_runtime": 1.0462,
|
| 561 |
+
"eval_samples_per_second": 574.434,
|
| 562 |
+
"eval_steps_per_second": 36.32,
|
| 563 |
+
"eval_weighted_f1": 0.8548245021423386,
|
| 564 |
+
"step": 1647
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 9.016460905349794,
|
| 568 |
+
"grad_norm": 148.36126708984375,
|
| 569 |
+
"learning_rate": 1.2210078931390409e-05,
|
| 570 |
+
"loss": 0.4445,
|
| 571 |
+
"step": 1650
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 9.15363511659808,
|
| 575 |
+
"grad_norm": 2.724369764328003,
|
| 576 |
+
"learning_rate": 1.2058287795992716e-05,
|
| 577 |
+
"loss": 0.255,
|
| 578 |
+
"step": 1675
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 9.290809327846365,
|
| 582 |
+
"grad_norm": 0.5079428553581238,
|
| 583 |
+
"learning_rate": 1.1906496660595022e-05,
|
| 584 |
+
"loss": 0.3072,
|
| 585 |
+
"step": 1700
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 9.427983539094651,
|
| 589 |
+
"grad_norm": 1.7856823205947876,
|
| 590 |
+
"learning_rate": 1.1754705525197329e-05,
|
| 591 |
+
"loss": 0.3519,
|
| 592 |
+
"step": 1725
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 9.565157750342935,
|
| 596 |
+
"grad_norm": 60.1046257019043,
|
| 597 |
+
"learning_rate": 1.1602914389799638e-05,
|
| 598 |
+
"loss": 0.2405,
|
| 599 |
+
"step": 1750
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 9.70233196159122,
|
| 603 |
+
"grad_norm": 226.64288330078125,
|
| 604 |
+
"learning_rate": 1.1451123254401944e-05,
|
| 605 |
+
"loss": 0.2963,
|
| 606 |
+
"step": 1775
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 9.839506172839506,
|
| 610 |
+
"grad_norm": 0.20350158214569092,
|
| 611 |
+
"learning_rate": 1.1299332119004251e-05,
|
| 612 |
+
"loss": 0.2051,
|
| 613 |
+
"step": 1800
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 9.976680384087791,
|
| 617 |
+
"grad_norm": 74.67986297607422,
|
| 618 |
+
"learning_rate": 1.1147540983606557e-05,
|
| 619 |
+
"loss": 0.2077,
|
| 620 |
+
"step": 1825
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 10.0,
|
| 624 |
+
"eval_accuracy": 0.848585690515807,
|
| 625 |
+
"eval_loss": 0.9442906379699707,
|
| 626 |
+
"eval_macro_f1": 0.8482438489142932,
|
| 627 |
+
"eval_runtime": 1.0755,
|
| 628 |
+
"eval_samples_per_second": 558.824,
|
| 629 |
+
"eval_steps_per_second": 35.333,
|
| 630 |
+
"eval_weighted_f1": 0.8481881734227498,
|
| 631 |
+
"step": 1830
|
| 632 |
+
},
|
| 633 |
+
{
|
| 634 |
+
"epoch": 10.109739368998628,
|
| 635 |
+
"grad_norm": 6.473784923553467,
|
| 636 |
+
"learning_rate": 1.0995749848208866e-05,
|
| 637 |
+
"loss": 0.0584,
|
| 638 |
+
"step": 1850
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 10.246913580246913,
|
| 642 |
+
"grad_norm": 4.44045352935791,
|
| 643 |
+
"learning_rate": 1.0843958712811173e-05,
|
| 644 |
+
"loss": 0.2697,
|
| 645 |
+
"step": 1875
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 10.3840877914952,
|
| 649 |
+
"grad_norm": 0.07267064601182938,
|
| 650 |
+
"learning_rate": 1.0692167577413479e-05,
|
| 651 |
+
"loss": 0.0803,
|
| 652 |
+
"step": 1900
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"epoch": 10.521262002743484,
|
| 656 |
+
"grad_norm": 0.06884893774986267,
|
| 657 |
+
"learning_rate": 1.0540376442015788e-05,
|
| 658 |
+
"loss": 0.2172,
|
| 659 |
+
"step": 1925
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 10.65843621399177,
|
| 663 |
+
"grad_norm": 98.49237060546875,
|
| 664 |
+
"learning_rate": 1.0388585306618095e-05,
|
| 665 |
+
"loss": 0.2178,
|
| 666 |
+
"step": 1950
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 10.795610425240055,
|
| 670 |
+
"grad_norm": 0.14222553372383118,
|
| 671 |
+
"learning_rate": 1.02367941712204e-05,
|
| 672 |
+
"loss": 0.1371,
|
| 673 |
+
"step": 1975
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 10.932784636488341,
|
| 677 |
+
"grad_norm": 0.06033104285597801,
|
| 678 |
+
"learning_rate": 1.008500303582271e-05,
|
| 679 |
+
"loss": 0.1399,
|
| 680 |
+
"step": 2000
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 11.0,
|
| 684 |
+
"eval_accuracy": 0.8519134775374376,
|
| 685 |
+
"eval_loss": 0.9461697340011597,
|
| 686 |
+
"eval_macro_f1": 0.8520281953984341,
|
| 687 |
+
"eval_runtime": 1.0477,
|
| 688 |
+
"eval_samples_per_second": 573.659,
|
| 689 |
+
"eval_steps_per_second": 36.271,
|
| 690 |
+
"eval_weighted_f1": 0.8518936768929636,
|
| 691 |
+
"step": 2013
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 11.065843621399177,
|
| 695 |
+
"grad_norm": 4.720425605773926,
|
| 696 |
+
"learning_rate": 9.933211900425015e-06,
|
| 697 |
+
"loss": 0.1719,
|
| 698 |
+
"step": 2025
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 11.203017832647463,
|
| 702 |
+
"grad_norm": 0.07295724004507065,
|
| 703 |
+
"learning_rate": 9.781420765027323e-06,
|
| 704 |
+
"loss": 0.1284,
|
| 705 |
+
"step": 2050
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 11.340192043895748,
|
| 709 |
+
"grad_norm": 4.621620178222656,
|
| 710 |
+
"learning_rate": 9.62962962962963e-06,
|
| 711 |
+
"loss": 0.2236,
|
| 712 |
+
"step": 2075
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 11.477366255144032,
|
| 716 |
+
"grad_norm": 0.08362743258476257,
|
| 717 |
+
"learning_rate": 9.477838494231937e-06,
|
| 718 |
+
"loss": 0.2323,
|
| 719 |
+
"step": 2100
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 11.614540466392318,
|
| 723 |
+
"grad_norm": 0.046199969947338104,
|
| 724 |
+
"learning_rate": 9.326047358834245e-06,
|
| 725 |
+
"loss": 0.0466,
|
| 726 |
+
"step": 2125
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 11.751714677640603,
|
| 730 |
+
"grad_norm": 0.10668723285198212,
|
| 731 |
+
"learning_rate": 9.174256223436552e-06,
|
| 732 |
+
"loss": 0.1513,
|
| 733 |
+
"step": 2150
|
| 734 |
+
},
|
| 735 |
+
{
|
| 736 |
+
"epoch": 11.88888888888889,
|
| 737 |
+
"grad_norm": 171.67169189453125,
|
| 738 |
+
"learning_rate": 9.02246508803886e-06,
|
| 739 |
+
"loss": 0.2603,
|
| 740 |
+
"step": 2175
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 12.0,
|
| 744 |
+
"eval_accuracy": 0.8452579034941764,
|
| 745 |
+
"eval_loss": 1.0590239763259888,
|
| 746 |
+
"eval_macro_f1": 0.8452238708770775,
|
| 747 |
+
"eval_runtime": 1.0353,
|
| 748 |
+
"eval_samples_per_second": 580.481,
|
| 749 |
+
"eval_steps_per_second": 36.703,
|
| 750 |
+
"eval_weighted_f1": 0.8449100720572259,
|
| 751 |
+
"step": 2196
|
| 752 |
+
}
|
| 753 |
+
],
|
| 754 |
+
"logging_steps": 25,
|
| 755 |
+
"max_steps": 3660,
|
| 756 |
+
"num_input_tokens_seen": 0,
|
| 757 |
+
"num_train_epochs": 20,
|
| 758 |
+
"save_steps": 500,
|
| 759 |
+
"stateful_callbacks": {
|
| 760 |
+
"EarlyStoppingCallback": {
|
| 761 |
+
"args": {
|
| 762 |
+
"early_stopping_patience": 3,
|
| 763 |
+
"early_stopping_threshold": 0.0
|
| 764 |
+
},
|
| 765 |
+
"attributes": {
|
| 766 |
+
"early_stopping_patience_counter": 3
|
| 767 |
+
}
|
| 768 |
+
},
|
| 769 |
+
"TrainerControl": {
|
| 770 |
+
"args": {
|
| 771 |
+
"should_epoch_stop": false,
|
| 772 |
+
"should_evaluate": false,
|
| 773 |
+
"should_log": false,
|
| 774 |
+
"should_save": true,
|
| 775 |
+
"should_training_stop": true
|
| 776 |
+
},
|
| 777 |
+
"attributes": {}
|
| 778 |
+
}
|
| 779 |
+
},
|
| 780 |
+
"total_flos": 0.0,
|
| 781 |
+
"train_batch_size": 8,
|
| 782 |
+
"trial_name": null,
|
| 783 |
+
"trial_params": null
|
| 784 |
+
}
|
ood-vifactcheck-bamibert-source-source-04-s42-20260828T233102Z-6235c159/trainer/checkpoint-1760/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-source-source-04-s42-20260828T233102Z-6235c159/trainer/checkpoint-1760/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-vifactcheck-bamibert-source-source-04-s42-20260828T233102Z-6235c159/trainer/checkpoint-1760/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-source-source-04-s42-20260828T233102Z-6235c159/trainer/checkpoint-1760/trainer_state.json
ADDED
|
@@ -0,0 +1,643 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1760,
|
| 3 |
+
"best_metric": 0.8546697114483482,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-source-source-04-s42-20260828T233102Z-6235c159/trainer/checkpoint-1760",
|
| 5 |
+
"epoch": 10.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1760,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.14245014245014245,
|
| 14 |
+
"grad_norm": 9.394332885742188,
|
| 15 |
+
"learning_rate": 1.3636363636363636e-06,
|
| 16 |
+
"loss": 4.4569,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.2849002849002849,
|
| 21 |
+
"grad_norm": 9.782153129577637,
|
| 22 |
+
"learning_rate": 2.784090909090909e-06,
|
| 23 |
+
"loss": 4.3965,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.42735042735042733,
|
| 28 |
+
"grad_norm": 10.704619407653809,
|
| 29 |
+
"learning_rate": 4.204545454545455e-06,
|
| 30 |
+
"loss": 4.4534,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5698005698005698,
|
| 35 |
+
"grad_norm": 8.475692749023438,
|
| 36 |
+
"learning_rate": 5.625e-06,
|
| 37 |
+
"loss": 4.4012,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.7122507122507122,
|
| 42 |
+
"grad_norm": 164.22633361816406,
|
| 43 |
+
"learning_rate": 7.045454545454546e-06,
|
| 44 |
+
"loss": 4.3972,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.8547008547008547,
|
| 49 |
+
"grad_norm": 11.407520294189453,
|
| 50 |
+
"learning_rate": 8.465909090909091e-06,
|
| 51 |
+
"loss": 4.3556,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.9971509971509972,
|
| 56 |
+
"grad_norm": 29.401058197021484,
|
| 57 |
+
"learning_rate": 9.886363636363637e-06,
|
| 58 |
+
"loss": 4.3456,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 1.0,
|
| 63 |
+
"eval_accuracy": 0.38461538461538464,
|
| 64 |
+
"eval_loss": 1.0725972652435303,
|
| 65 |
+
"eval_macro_f1": 0.286802219720175,
|
| 66 |
+
"eval_runtime": 1.5196,
|
| 67 |
+
"eval_samples_per_second": 453.416,
|
| 68 |
+
"eval_steps_per_second": 28.955,
|
| 69 |
+
"eval_weighted_f1": 0.28611128191860047,
|
| 70 |
+
"step": 176
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.1367521367521367,
|
| 74 |
+
"grad_norm": 24.720022201538086,
|
| 75 |
+
"learning_rate": 1.1306818181818183e-05,
|
| 76 |
+
"loss": 4.1618,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.2792022792022792,
|
| 81 |
+
"grad_norm": 37.82431411743164,
|
| 82 |
+
"learning_rate": 1.2727272727272728e-05,
|
| 83 |
+
"loss": 4.1084,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.4216524216524218,
|
| 88 |
+
"grad_norm": 50.1373291015625,
|
| 89 |
+
"learning_rate": 1.4147727272727274e-05,
|
| 90 |
+
"loss": 3.8109,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.564102564102564,
|
| 95 |
+
"grad_norm": 40.463191986083984,
|
| 96 |
+
"learning_rate": 1.5568181818181822e-05,
|
| 97 |
+
"loss": 3.1555,
|
| 98 |
+
"step": 275
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.7065527065527064,
|
| 102 |
+
"grad_norm": 79.27658081054688,
|
| 103 |
+
"learning_rate": 1.6988636363636368e-05,
|
| 104 |
+
"loss": 2.848,
|
| 105 |
+
"step": 300
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.849002849002849,
|
| 109 |
+
"grad_norm": 49.66876220703125,
|
| 110 |
+
"learning_rate": 1.840909090909091e-05,
|
| 111 |
+
"loss": 2.9216,
|
| 112 |
+
"step": 325
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 1.9914529914529915,
|
| 116 |
+
"grad_norm": 36.2674674987793,
|
| 117 |
+
"learning_rate": 1.9829545454545456e-05,
|
| 118 |
+
"loss": 2.5689,
|
| 119 |
+
"step": 350
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 2.0,
|
| 123 |
+
"eval_accuracy": 0.7692307692307693,
|
| 124 |
+
"eval_loss": 0.610457181930542,
|
| 125 |
+
"eval_macro_f1": 0.7560360879833737,
|
| 126 |
+
"eval_runtime": 1.4543,
|
| 127 |
+
"eval_samples_per_second": 473.752,
|
| 128 |
+
"eval_steps_per_second": 30.254,
|
| 129 |
+
"eval_weighted_f1": 0.7573535366234558,
|
| 130 |
+
"step": 352
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.131054131054131,
|
| 134 |
+
"grad_norm": 45.669944763183594,
|
| 135 |
+
"learning_rate": 1.9861111111111114e-05,
|
| 136 |
+
"loss": 2.2152,
|
| 137 |
+
"step": 375
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 2.2735042735042734,
|
| 141 |
+
"grad_norm": 66.53202056884766,
|
| 142 |
+
"learning_rate": 1.9703282828282832e-05,
|
| 143 |
+
"loss": 2.2855,
|
| 144 |
+
"step": 400
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.415954415954416,
|
| 148 |
+
"grad_norm": 39.979576110839844,
|
| 149 |
+
"learning_rate": 1.9545454545454546e-05,
|
| 150 |
+
"loss": 2.0533,
|
| 151 |
+
"step": 425
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.5584045584045585,
|
| 155 |
+
"grad_norm": 43.54996109008789,
|
| 156 |
+
"learning_rate": 1.9387626262626263e-05,
|
| 157 |
+
"loss": 1.8676,
|
| 158 |
+
"step": 450
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.700854700854701,
|
| 162 |
+
"grad_norm": 66.63773345947266,
|
| 163 |
+
"learning_rate": 1.922979797979798e-05,
|
| 164 |
+
"loss": 2.0017,
|
| 165 |
+
"step": 475
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.8433048433048436,
|
| 169 |
+
"grad_norm": 63.00020980834961,
|
| 170 |
+
"learning_rate": 1.90719696969697e-05,
|
| 171 |
+
"loss": 2.0362,
|
| 172 |
+
"step": 500
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 2.9857549857549857,
|
| 176 |
+
"grad_norm": 35.12709045410156,
|
| 177 |
+
"learning_rate": 1.8914141414141416e-05,
|
| 178 |
+
"loss": 1.8993,
|
| 179 |
+
"step": 525
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 3.0,
|
| 183 |
+
"eval_accuracy": 0.8214804063860668,
|
| 184 |
+
"eval_loss": 0.4591293931007385,
|
| 185 |
+
"eval_macro_f1": 0.8187462106799788,
|
| 186 |
+
"eval_runtime": 1.475,
|
| 187 |
+
"eval_samples_per_second": 467.133,
|
| 188 |
+
"eval_steps_per_second": 29.831,
|
| 189 |
+
"eval_weighted_f1": 0.8189515251053842,
|
| 190 |
+
"step": 528
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 3.1253561253561255,
|
| 194 |
+
"grad_norm": 39.91587829589844,
|
| 195 |
+
"learning_rate": 1.8756313131313133e-05,
|
| 196 |
+
"loss": 1.4697,
|
| 197 |
+
"step": 550
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 3.267806267806268,
|
| 201 |
+
"grad_norm": 53.09515380859375,
|
| 202 |
+
"learning_rate": 1.859848484848485e-05,
|
| 203 |
+
"loss": 1.2339,
|
| 204 |
+
"step": 575
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.41025641025641,
|
| 208 |
+
"grad_norm": 27.454301834106445,
|
| 209 |
+
"learning_rate": 1.844065656565657e-05,
|
| 210 |
+
"loss": 1.3934,
|
| 211 |
+
"step": 600
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.5527065527065527,
|
| 215 |
+
"grad_norm": 30.73740577697754,
|
| 216 |
+
"learning_rate": 1.8282828282828286e-05,
|
| 217 |
+
"loss": 1.5034,
|
| 218 |
+
"step": 625
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.695156695156695,
|
| 222 |
+
"grad_norm": 23.839885711669922,
|
| 223 |
+
"learning_rate": 1.8125e-05,
|
| 224 |
+
"loss": 1.2803,
|
| 225 |
+
"step": 650
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.8376068376068377,
|
| 229 |
+
"grad_norm": 29.481029510498047,
|
| 230 |
+
"learning_rate": 1.7967171717171718e-05,
|
| 231 |
+
"loss": 1.401,
|
| 232 |
+
"step": 675
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.98005698005698,
|
| 236 |
+
"grad_norm": 86.43328857421875,
|
| 237 |
+
"learning_rate": 1.7809343434343435e-05,
|
| 238 |
+
"loss": 1.4397,
|
| 239 |
+
"step": 700
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 4.0,
|
| 243 |
+
"eval_accuracy": 0.8258345428156749,
|
| 244 |
+
"eval_loss": 0.5382989645004272,
|
| 245 |
+
"eval_macro_f1": 0.8243082549738691,
|
| 246 |
+
"eval_runtime": 1.5795,
|
| 247 |
+
"eval_samples_per_second": 436.224,
|
| 248 |
+
"eval_steps_per_second": 27.858,
|
| 249 |
+
"eval_weighted_f1": 0.8242691978542168,
|
| 250 |
+
"step": 704
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 4.119658119658119,
|
| 254 |
+
"grad_norm": 53.799198150634766,
|
| 255 |
+
"learning_rate": 1.7651515151515153e-05,
|
| 256 |
+
"loss": 1.1143,
|
| 257 |
+
"step": 725
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 4.262108262108262,
|
| 261 |
+
"grad_norm": 31.68073844909668,
|
| 262 |
+
"learning_rate": 1.749368686868687e-05,
|
| 263 |
+
"loss": 1.0672,
|
| 264 |
+
"step": 750
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.404558404558404,
|
| 268 |
+
"grad_norm": 32.63469696044922,
|
| 269 |
+
"learning_rate": 1.7335858585858588e-05,
|
| 270 |
+
"loss": 0.9712,
|
| 271 |
+
"step": 775
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 4.547008547008547,
|
| 275 |
+
"grad_norm": 64.61741638183594,
|
| 276 |
+
"learning_rate": 1.7178030303030305e-05,
|
| 277 |
+
"loss": 0.9317,
|
| 278 |
+
"step": 800
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 4.689458689458689,
|
| 282 |
+
"grad_norm": 60.9797477722168,
|
| 283 |
+
"learning_rate": 1.7020202020202023e-05,
|
| 284 |
+
"loss": 1.045,
|
| 285 |
+
"step": 825
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 4.831908831908832,
|
| 289 |
+
"grad_norm": 98.00072479248047,
|
| 290 |
+
"learning_rate": 1.6862373737373737e-05,
|
| 291 |
+
"loss": 0.7834,
|
| 292 |
+
"step": 850
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 4.9743589743589745,
|
| 296 |
+
"grad_norm": 100.28450012207031,
|
| 297 |
+
"learning_rate": 1.6704545454545454e-05,
|
| 298 |
+
"loss": 0.9139,
|
| 299 |
+
"step": 875
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 5.0,
|
| 303 |
+
"eval_accuracy": 0.8330914368650217,
|
| 304 |
+
"eval_loss": 0.5526773929595947,
|
| 305 |
+
"eval_macro_f1": 0.8346436369719524,
|
| 306 |
+
"eval_runtime": 1.4524,
|
| 307 |
+
"eval_samples_per_second": 474.401,
|
| 308 |
+
"eval_steps_per_second": 30.296,
|
| 309 |
+
"eval_weighted_f1": 0.8349771026424844,
|
| 310 |
+
"step": 880
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 5.113960113960114,
|
| 314 |
+
"grad_norm": 61.7004508972168,
|
| 315 |
+
"learning_rate": 1.6546717171717172e-05,
|
| 316 |
+
"loss": 0.7074,
|
| 317 |
+
"step": 900
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 5.256410256410256,
|
| 321 |
+
"grad_norm": 42.83418273925781,
|
| 322 |
+
"learning_rate": 1.638888888888889e-05,
|
| 323 |
+
"loss": 0.687,
|
| 324 |
+
"step": 925
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 5.398860398860399,
|
| 328 |
+
"grad_norm": 39.12403869628906,
|
| 329 |
+
"learning_rate": 1.6231060606060607e-05,
|
| 330 |
+
"loss": 0.4452,
|
| 331 |
+
"step": 950
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.5413105413105415,
|
| 335 |
+
"grad_norm": 38.58466339111328,
|
| 336 |
+
"learning_rate": 1.6073232323232324e-05,
|
| 337 |
+
"loss": 0.5196,
|
| 338 |
+
"step": 975
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 5.683760683760684,
|
| 342 |
+
"grad_norm": 32.99896240234375,
|
| 343 |
+
"learning_rate": 1.5915404040404042e-05,
|
| 344 |
+
"loss": 0.7703,
|
| 345 |
+
"step": 1000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.8262108262108265,
|
| 349 |
+
"grad_norm": 70.25640106201172,
|
| 350 |
+
"learning_rate": 1.575757575757576e-05,
|
| 351 |
+
"loss": 0.7204,
|
| 352 |
+
"step": 1025
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 5.968660968660968,
|
| 356 |
+
"grad_norm": 28.564496994018555,
|
| 357 |
+
"learning_rate": 1.5599747474747477e-05,
|
| 358 |
+
"loss": 0.6164,
|
| 359 |
+
"step": 1050
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 6.0,
|
| 363 |
+
"eval_accuracy": 0.8388969521044993,
|
| 364 |
+
"eval_loss": 0.6260063648223877,
|
| 365 |
+
"eval_macro_f1": 0.8365539730815194,
|
| 366 |
+
"eval_runtime": 1.5255,
|
| 367 |
+
"eval_samples_per_second": 451.641,
|
| 368 |
+
"eval_steps_per_second": 28.842,
|
| 369 |
+
"eval_weighted_f1": 0.8370927812704533,
|
| 370 |
+
"step": 1056
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 6.1082621082621085,
|
| 374 |
+
"grad_norm": 31.06218719482422,
|
| 375 |
+
"learning_rate": 1.544191919191919e-05,
|
| 376 |
+
"loss": 0.4532,
|
| 377 |
+
"step": 1075
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 6.250712250712251,
|
| 381 |
+
"grad_norm": 35.933807373046875,
|
| 382 |
+
"learning_rate": 1.528409090909091e-05,
|
| 383 |
+
"loss": 0.4479,
|
| 384 |
+
"step": 1100
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 6.3931623931623935,
|
| 388 |
+
"grad_norm": 65.59700012207031,
|
| 389 |
+
"learning_rate": 1.5126262626262626e-05,
|
| 390 |
+
"loss": 0.5564,
|
| 391 |
+
"step": 1125
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 6.535612535612536,
|
| 395 |
+
"grad_norm": 7.167136192321777,
|
| 396 |
+
"learning_rate": 1.4968434343434344e-05,
|
| 397 |
+
"loss": 0.5108,
|
| 398 |
+
"step": 1150
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 6.678062678062678,
|
| 402 |
+
"grad_norm": 147.64076232910156,
|
| 403 |
+
"learning_rate": 1.4810606060606061e-05,
|
| 404 |
+
"loss": 0.4343,
|
| 405 |
+
"step": 1175
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 6.82051282051282,
|
| 409 |
+
"grad_norm": 2.47883677482605,
|
| 410 |
+
"learning_rate": 1.4652777777777779e-05,
|
| 411 |
+
"loss": 0.6561,
|
| 412 |
+
"step": 1200
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 6.962962962962963,
|
| 416 |
+
"grad_norm": 52.5629768371582,
|
| 417 |
+
"learning_rate": 1.4494949494949494e-05,
|
| 418 |
+
"loss": 0.6477,
|
| 419 |
+
"step": 1225
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 7.0,
|
| 423 |
+
"eval_accuracy": 0.8490566037735849,
|
| 424 |
+
"eval_loss": 0.6626464128494263,
|
| 425 |
+
"eval_macro_f1": 0.8480730502037321,
|
| 426 |
+
"eval_runtime": 1.4391,
|
| 427 |
+
"eval_samples_per_second": 478.764,
|
| 428 |
+
"eval_steps_per_second": 30.574,
|
| 429 |
+
"eval_weighted_f1": 0.848201027008074,
|
| 430 |
+
"step": 1232
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 7.102564102564102,
|
| 434 |
+
"grad_norm": 9.221817016601562,
|
| 435 |
+
"learning_rate": 1.4337121212121212e-05,
|
| 436 |
+
"loss": 0.3944,
|
| 437 |
+
"step": 1250
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 7.245014245014245,
|
| 441 |
+
"grad_norm": 13.65827465057373,
|
| 442 |
+
"learning_rate": 1.4179292929292931e-05,
|
| 443 |
+
"loss": 0.3707,
|
| 444 |
+
"step": 1275
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 7.387464387464387,
|
| 448 |
+
"grad_norm": 102.12430572509766,
|
| 449 |
+
"learning_rate": 1.4021464646464649e-05,
|
| 450 |
+
"loss": 0.3838,
|
| 451 |
+
"step": 1300
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 7.52991452991453,
|
| 455 |
+
"grad_norm": 41.775047302246094,
|
| 456 |
+
"learning_rate": 1.3863636363636366e-05,
|
| 457 |
+
"loss": 0.3613,
|
| 458 |
+
"step": 1325
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 7.672364672364672,
|
| 462 |
+
"grad_norm": 7.700901985168457,
|
| 463 |
+
"learning_rate": 1.3705808080808084e-05,
|
| 464 |
+
"loss": 0.3836,
|
| 465 |
+
"step": 1350
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 7.814814814814815,
|
| 469 |
+
"grad_norm": 30.64459800720215,
|
| 470 |
+
"learning_rate": 1.35479797979798e-05,
|
| 471 |
+
"loss": 0.4086,
|
| 472 |
+
"step": 1375
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 7.957264957264957,
|
| 476 |
+
"grad_norm": 84.40220642089844,
|
| 477 |
+
"learning_rate": 1.3390151515151517e-05,
|
| 478 |
+
"loss": 0.348,
|
| 479 |
+
"step": 1400
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 8.0,
|
| 483 |
+
"eval_accuracy": 0.8403483309143687,
|
| 484 |
+
"eval_loss": 0.8505121469497681,
|
| 485 |
+
"eval_macro_f1": 0.837636893114618,
|
| 486 |
+
"eval_runtime": 1.5391,
|
| 487 |
+
"eval_samples_per_second": 447.668,
|
| 488 |
+
"eval_steps_per_second": 28.588,
|
| 489 |
+
"eval_weighted_f1": 0.8379419870932212,
|
| 490 |
+
"step": 1408
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 8.096866096866098,
|
| 494 |
+
"grad_norm": 7.839972496032715,
|
| 495 |
+
"learning_rate": 1.3232323232323234e-05,
|
| 496 |
+
"loss": 0.1529,
|
| 497 |
+
"step": 1425
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 8.239316239316238,
|
| 501 |
+
"grad_norm": 147.12326049804688,
|
| 502 |
+
"learning_rate": 1.3074494949494952e-05,
|
| 503 |
+
"loss": 0.2763,
|
| 504 |
+
"step": 1450
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 8.381766381766381,
|
| 508 |
+
"grad_norm": 21.040546417236328,
|
| 509 |
+
"learning_rate": 1.2916666666666668e-05,
|
| 510 |
+
"loss": 0.3735,
|
| 511 |
+
"step": 1475
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 8.524216524216524,
|
| 515 |
+
"grad_norm": 82.0713882446289,
|
| 516 |
+
"learning_rate": 1.2758838383838385e-05,
|
| 517 |
+
"loss": 0.427,
|
| 518 |
+
"step": 1500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 8.666666666666666,
|
| 522 |
+
"grad_norm": 62.74919509887695,
|
| 523 |
+
"learning_rate": 1.2601010101010103e-05,
|
| 524 |
+
"loss": 0.2118,
|
| 525 |
+
"step": 1525
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 8.809116809116809,
|
| 529 |
+
"grad_norm": 0.20923995971679688,
|
| 530 |
+
"learning_rate": 1.244318181818182e-05,
|
| 531 |
+
"loss": 0.5143,
|
| 532 |
+
"step": 1550
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 8.951566951566951,
|
| 536 |
+
"grad_norm": 0.7549470663070679,
|
| 537 |
+
"learning_rate": 1.2285353535353536e-05,
|
| 538 |
+
"loss": 0.4299,
|
| 539 |
+
"step": 1575
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 9.0,
|
| 543 |
+
"eval_accuracy": 0.8258345428156749,
|
| 544 |
+
"eval_loss": 1.0375233888626099,
|
| 545 |
+
"eval_macro_f1": 0.8240438177651709,
|
| 546 |
+
"eval_runtime": 1.4931,
|
| 547 |
+
"eval_samples_per_second": 461.46,
|
| 548 |
+
"eval_steps_per_second": 29.469,
|
| 549 |
+
"eval_weighted_f1": 0.8240742430839567,
|
| 550 |
+
"step": 1584
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 9.091168091168091,
|
| 554 |
+
"grad_norm": 6.6440863609313965,
|
| 555 |
+
"learning_rate": 1.2127525252525254e-05,
|
| 556 |
+
"loss": 0.2515,
|
| 557 |
+
"step": 1600
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 9.233618233618234,
|
| 561 |
+
"grad_norm": 156.8964385986328,
|
| 562 |
+
"learning_rate": 1.1969696969696971e-05,
|
| 563 |
+
"loss": 0.2933,
|
| 564 |
+
"step": 1625
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 9.376068376068377,
|
| 568 |
+
"grad_norm": 10.411484718322754,
|
| 569 |
+
"learning_rate": 1.1811868686868689e-05,
|
| 570 |
+
"loss": 0.3251,
|
| 571 |
+
"step": 1650
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 9.518518518518519,
|
| 575 |
+
"grad_norm": 8.135589599609375,
|
| 576 |
+
"learning_rate": 1.1654040404040406e-05,
|
| 577 |
+
"loss": 0.1041,
|
| 578 |
+
"step": 1675
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 9.660968660968662,
|
| 582 |
+
"grad_norm": 236.67787170410156,
|
| 583 |
+
"learning_rate": 1.1496212121212122e-05,
|
| 584 |
+
"loss": 0.2226,
|
| 585 |
+
"step": 1700
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 9.803418803418804,
|
| 589 |
+
"grad_norm": 97.63804626464844,
|
| 590 |
+
"learning_rate": 1.133838383838384e-05,
|
| 591 |
+
"loss": 0.234,
|
| 592 |
+
"step": 1725
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 9.945868945868947,
|
| 596 |
+
"grad_norm": 0.3027924597263336,
|
| 597 |
+
"learning_rate": 1.1180555555555557e-05,
|
| 598 |
+
"loss": 0.2665,
|
| 599 |
+
"step": 1750
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 10.0,
|
| 603 |
+
"eval_accuracy": 0.8548621190130624,
|
| 604 |
+
"eval_loss": 0.9538707137107849,
|
| 605 |
+
"eval_macro_f1": 0.8546697114483482,
|
| 606 |
+
"eval_runtime": 1.5132,
|
| 607 |
+
"eval_samples_per_second": 455.327,
|
| 608 |
+
"eval_steps_per_second": 29.078,
|
| 609 |
+
"eval_weighted_f1": 0.8548730047689096,
|
| 610 |
+
"step": 1760
|
| 611 |
+
}
|
| 612 |
+
],
|
| 613 |
+
"logging_steps": 25,
|
| 614 |
+
"max_steps": 3520,
|
| 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": 0
|
| 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": false
|
| 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-source-source-04-s42-20260828T233102Z-6235c159/trainer/checkpoint-2288/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-source-source-04-s42-20260828T233102Z-6235c159/trainer/checkpoint-2288/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-vifactcheck-bamibert-source-source-04-s42-20260828T233102Z-6235c159/trainer/checkpoint-2288/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-source-source-04-s42-20260828T233102Z-6235c159/trainer/checkpoint-2288/trainer_state.json
ADDED
|
@@ -0,0 +1,823 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1760,
|
| 3 |
+
"best_metric": 0.8546697114483482,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-source-source-04-s42-20260828T233102Z-6235c159/trainer/checkpoint-1760",
|
| 5 |
+
"epoch": 13.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 2288,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.14245014245014245,
|
| 14 |
+
"grad_norm": 9.394332885742188,
|
| 15 |
+
"learning_rate": 1.3636363636363636e-06,
|
| 16 |
+
"loss": 4.4569,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.2849002849002849,
|
| 21 |
+
"grad_norm": 9.782153129577637,
|
| 22 |
+
"learning_rate": 2.784090909090909e-06,
|
| 23 |
+
"loss": 4.3965,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.42735042735042733,
|
| 28 |
+
"grad_norm": 10.704619407653809,
|
| 29 |
+
"learning_rate": 4.204545454545455e-06,
|
| 30 |
+
"loss": 4.4534,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5698005698005698,
|
| 35 |
+
"grad_norm": 8.475692749023438,
|
| 36 |
+
"learning_rate": 5.625e-06,
|
| 37 |
+
"loss": 4.4012,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.7122507122507122,
|
| 42 |
+
"grad_norm": 164.22633361816406,
|
| 43 |
+
"learning_rate": 7.045454545454546e-06,
|
| 44 |
+
"loss": 4.3972,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.8547008547008547,
|
| 49 |
+
"grad_norm": 11.407520294189453,
|
| 50 |
+
"learning_rate": 8.465909090909091e-06,
|
| 51 |
+
"loss": 4.3556,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.9971509971509972,
|
| 56 |
+
"grad_norm": 29.401058197021484,
|
| 57 |
+
"learning_rate": 9.886363636363637e-06,
|
| 58 |
+
"loss": 4.3456,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 1.0,
|
| 63 |
+
"eval_accuracy": 0.38461538461538464,
|
| 64 |
+
"eval_loss": 1.0725972652435303,
|
| 65 |
+
"eval_macro_f1": 0.286802219720175,
|
| 66 |
+
"eval_runtime": 1.5196,
|
| 67 |
+
"eval_samples_per_second": 453.416,
|
| 68 |
+
"eval_steps_per_second": 28.955,
|
| 69 |
+
"eval_weighted_f1": 0.28611128191860047,
|
| 70 |
+
"step": 176
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.1367521367521367,
|
| 74 |
+
"grad_norm": 24.720022201538086,
|
| 75 |
+
"learning_rate": 1.1306818181818183e-05,
|
| 76 |
+
"loss": 4.1618,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.2792022792022792,
|
| 81 |
+
"grad_norm": 37.82431411743164,
|
| 82 |
+
"learning_rate": 1.2727272727272728e-05,
|
| 83 |
+
"loss": 4.1084,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.4216524216524218,
|
| 88 |
+
"grad_norm": 50.1373291015625,
|
| 89 |
+
"learning_rate": 1.4147727272727274e-05,
|
| 90 |
+
"loss": 3.8109,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.564102564102564,
|
| 95 |
+
"grad_norm": 40.463191986083984,
|
| 96 |
+
"learning_rate": 1.5568181818181822e-05,
|
| 97 |
+
"loss": 3.1555,
|
| 98 |
+
"step": 275
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.7065527065527064,
|
| 102 |
+
"grad_norm": 79.27658081054688,
|
| 103 |
+
"learning_rate": 1.6988636363636368e-05,
|
| 104 |
+
"loss": 2.848,
|
| 105 |
+
"step": 300
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.849002849002849,
|
| 109 |
+
"grad_norm": 49.66876220703125,
|
| 110 |
+
"learning_rate": 1.840909090909091e-05,
|
| 111 |
+
"loss": 2.9216,
|
| 112 |
+
"step": 325
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 1.9914529914529915,
|
| 116 |
+
"grad_norm": 36.2674674987793,
|
| 117 |
+
"learning_rate": 1.9829545454545456e-05,
|
| 118 |
+
"loss": 2.5689,
|
| 119 |
+
"step": 350
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 2.0,
|
| 123 |
+
"eval_accuracy": 0.7692307692307693,
|
| 124 |
+
"eval_loss": 0.610457181930542,
|
| 125 |
+
"eval_macro_f1": 0.7560360879833737,
|
| 126 |
+
"eval_runtime": 1.4543,
|
| 127 |
+
"eval_samples_per_second": 473.752,
|
| 128 |
+
"eval_steps_per_second": 30.254,
|
| 129 |
+
"eval_weighted_f1": 0.7573535366234558,
|
| 130 |
+
"step": 352
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.131054131054131,
|
| 134 |
+
"grad_norm": 45.669944763183594,
|
| 135 |
+
"learning_rate": 1.9861111111111114e-05,
|
| 136 |
+
"loss": 2.2152,
|
| 137 |
+
"step": 375
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 2.2735042735042734,
|
| 141 |
+
"grad_norm": 66.53202056884766,
|
| 142 |
+
"learning_rate": 1.9703282828282832e-05,
|
| 143 |
+
"loss": 2.2855,
|
| 144 |
+
"step": 400
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.415954415954416,
|
| 148 |
+
"grad_norm": 39.979576110839844,
|
| 149 |
+
"learning_rate": 1.9545454545454546e-05,
|
| 150 |
+
"loss": 2.0533,
|
| 151 |
+
"step": 425
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.5584045584045585,
|
| 155 |
+
"grad_norm": 43.54996109008789,
|
| 156 |
+
"learning_rate": 1.9387626262626263e-05,
|
| 157 |
+
"loss": 1.8676,
|
| 158 |
+
"step": 450
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.700854700854701,
|
| 162 |
+
"grad_norm": 66.63773345947266,
|
| 163 |
+
"learning_rate": 1.922979797979798e-05,
|
| 164 |
+
"loss": 2.0017,
|
| 165 |
+
"step": 475
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.8433048433048436,
|
| 169 |
+
"grad_norm": 63.00020980834961,
|
| 170 |
+
"learning_rate": 1.90719696969697e-05,
|
| 171 |
+
"loss": 2.0362,
|
| 172 |
+
"step": 500
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 2.9857549857549857,
|
| 176 |
+
"grad_norm": 35.12709045410156,
|
| 177 |
+
"learning_rate": 1.8914141414141416e-05,
|
| 178 |
+
"loss": 1.8993,
|
| 179 |
+
"step": 525
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 3.0,
|
| 183 |
+
"eval_accuracy": 0.8214804063860668,
|
| 184 |
+
"eval_loss": 0.4591293931007385,
|
| 185 |
+
"eval_macro_f1": 0.8187462106799788,
|
| 186 |
+
"eval_runtime": 1.475,
|
| 187 |
+
"eval_samples_per_second": 467.133,
|
| 188 |
+
"eval_steps_per_second": 29.831,
|
| 189 |
+
"eval_weighted_f1": 0.8189515251053842,
|
| 190 |
+
"step": 528
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 3.1253561253561255,
|
| 194 |
+
"grad_norm": 39.91587829589844,
|
| 195 |
+
"learning_rate": 1.8756313131313133e-05,
|
| 196 |
+
"loss": 1.4697,
|
| 197 |
+
"step": 550
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 3.267806267806268,
|
| 201 |
+
"grad_norm": 53.09515380859375,
|
| 202 |
+
"learning_rate": 1.859848484848485e-05,
|
| 203 |
+
"loss": 1.2339,
|
| 204 |
+
"step": 575
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.41025641025641,
|
| 208 |
+
"grad_norm": 27.454301834106445,
|
| 209 |
+
"learning_rate": 1.844065656565657e-05,
|
| 210 |
+
"loss": 1.3934,
|
| 211 |
+
"step": 600
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.5527065527065527,
|
| 215 |
+
"grad_norm": 30.73740577697754,
|
| 216 |
+
"learning_rate": 1.8282828282828286e-05,
|
| 217 |
+
"loss": 1.5034,
|
| 218 |
+
"step": 625
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.695156695156695,
|
| 222 |
+
"grad_norm": 23.839885711669922,
|
| 223 |
+
"learning_rate": 1.8125e-05,
|
| 224 |
+
"loss": 1.2803,
|
| 225 |
+
"step": 650
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.8376068376068377,
|
| 229 |
+
"grad_norm": 29.481029510498047,
|
| 230 |
+
"learning_rate": 1.7967171717171718e-05,
|
| 231 |
+
"loss": 1.401,
|
| 232 |
+
"step": 675
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.98005698005698,
|
| 236 |
+
"grad_norm": 86.43328857421875,
|
| 237 |
+
"learning_rate": 1.7809343434343435e-05,
|
| 238 |
+
"loss": 1.4397,
|
| 239 |
+
"step": 700
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 4.0,
|
| 243 |
+
"eval_accuracy": 0.8258345428156749,
|
| 244 |
+
"eval_loss": 0.5382989645004272,
|
| 245 |
+
"eval_macro_f1": 0.8243082549738691,
|
| 246 |
+
"eval_runtime": 1.5795,
|
| 247 |
+
"eval_samples_per_second": 436.224,
|
| 248 |
+
"eval_steps_per_second": 27.858,
|
| 249 |
+
"eval_weighted_f1": 0.8242691978542168,
|
| 250 |
+
"step": 704
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 4.119658119658119,
|
| 254 |
+
"grad_norm": 53.799198150634766,
|
| 255 |
+
"learning_rate": 1.7651515151515153e-05,
|
| 256 |
+
"loss": 1.1143,
|
| 257 |
+
"step": 725
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 4.262108262108262,
|
| 261 |
+
"grad_norm": 31.68073844909668,
|
| 262 |
+
"learning_rate": 1.749368686868687e-05,
|
| 263 |
+
"loss": 1.0672,
|
| 264 |
+
"step": 750
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.404558404558404,
|
| 268 |
+
"grad_norm": 32.63469696044922,
|
| 269 |
+
"learning_rate": 1.7335858585858588e-05,
|
| 270 |
+
"loss": 0.9712,
|
| 271 |
+
"step": 775
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 4.547008547008547,
|
| 275 |
+
"grad_norm": 64.61741638183594,
|
| 276 |
+
"learning_rate": 1.7178030303030305e-05,
|
| 277 |
+
"loss": 0.9317,
|
| 278 |
+
"step": 800
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 4.689458689458689,
|
| 282 |
+
"grad_norm": 60.9797477722168,
|
| 283 |
+
"learning_rate": 1.7020202020202023e-05,
|
| 284 |
+
"loss": 1.045,
|
| 285 |
+
"step": 825
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 4.831908831908832,
|
| 289 |
+
"grad_norm": 98.00072479248047,
|
| 290 |
+
"learning_rate": 1.6862373737373737e-05,
|
| 291 |
+
"loss": 0.7834,
|
| 292 |
+
"step": 850
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 4.9743589743589745,
|
| 296 |
+
"grad_norm": 100.28450012207031,
|
| 297 |
+
"learning_rate": 1.6704545454545454e-05,
|
| 298 |
+
"loss": 0.9139,
|
| 299 |
+
"step": 875
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 5.0,
|
| 303 |
+
"eval_accuracy": 0.8330914368650217,
|
| 304 |
+
"eval_loss": 0.5526773929595947,
|
| 305 |
+
"eval_macro_f1": 0.8346436369719524,
|
| 306 |
+
"eval_runtime": 1.4524,
|
| 307 |
+
"eval_samples_per_second": 474.401,
|
| 308 |
+
"eval_steps_per_second": 30.296,
|
| 309 |
+
"eval_weighted_f1": 0.8349771026424844,
|
| 310 |
+
"step": 880
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 5.113960113960114,
|
| 314 |
+
"grad_norm": 61.7004508972168,
|
| 315 |
+
"learning_rate": 1.6546717171717172e-05,
|
| 316 |
+
"loss": 0.7074,
|
| 317 |
+
"step": 900
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 5.256410256410256,
|
| 321 |
+
"grad_norm": 42.83418273925781,
|
| 322 |
+
"learning_rate": 1.638888888888889e-05,
|
| 323 |
+
"loss": 0.687,
|
| 324 |
+
"step": 925
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 5.398860398860399,
|
| 328 |
+
"grad_norm": 39.12403869628906,
|
| 329 |
+
"learning_rate": 1.6231060606060607e-05,
|
| 330 |
+
"loss": 0.4452,
|
| 331 |
+
"step": 950
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.5413105413105415,
|
| 335 |
+
"grad_norm": 38.58466339111328,
|
| 336 |
+
"learning_rate": 1.6073232323232324e-05,
|
| 337 |
+
"loss": 0.5196,
|
| 338 |
+
"step": 975
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 5.683760683760684,
|
| 342 |
+
"grad_norm": 32.99896240234375,
|
| 343 |
+
"learning_rate": 1.5915404040404042e-05,
|
| 344 |
+
"loss": 0.7703,
|
| 345 |
+
"step": 1000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.8262108262108265,
|
| 349 |
+
"grad_norm": 70.25640106201172,
|
| 350 |
+
"learning_rate": 1.575757575757576e-05,
|
| 351 |
+
"loss": 0.7204,
|
| 352 |
+
"step": 1025
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 5.968660968660968,
|
| 356 |
+
"grad_norm": 28.564496994018555,
|
| 357 |
+
"learning_rate": 1.5599747474747477e-05,
|
| 358 |
+
"loss": 0.6164,
|
| 359 |
+
"step": 1050
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 6.0,
|
| 363 |
+
"eval_accuracy": 0.8388969521044993,
|
| 364 |
+
"eval_loss": 0.6260063648223877,
|
| 365 |
+
"eval_macro_f1": 0.8365539730815194,
|
| 366 |
+
"eval_runtime": 1.5255,
|
| 367 |
+
"eval_samples_per_second": 451.641,
|
| 368 |
+
"eval_steps_per_second": 28.842,
|
| 369 |
+
"eval_weighted_f1": 0.8370927812704533,
|
| 370 |
+
"step": 1056
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 6.1082621082621085,
|
| 374 |
+
"grad_norm": 31.06218719482422,
|
| 375 |
+
"learning_rate": 1.544191919191919e-05,
|
| 376 |
+
"loss": 0.4532,
|
| 377 |
+
"step": 1075
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 6.250712250712251,
|
| 381 |
+
"grad_norm": 35.933807373046875,
|
| 382 |
+
"learning_rate": 1.528409090909091e-05,
|
| 383 |
+
"loss": 0.4479,
|
| 384 |
+
"step": 1100
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 6.3931623931623935,
|
| 388 |
+
"grad_norm": 65.59700012207031,
|
| 389 |
+
"learning_rate": 1.5126262626262626e-05,
|
| 390 |
+
"loss": 0.5564,
|
| 391 |
+
"step": 1125
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 6.535612535612536,
|
| 395 |
+
"grad_norm": 7.167136192321777,
|
| 396 |
+
"learning_rate": 1.4968434343434344e-05,
|
| 397 |
+
"loss": 0.5108,
|
| 398 |
+
"step": 1150
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 6.678062678062678,
|
| 402 |
+
"grad_norm": 147.64076232910156,
|
| 403 |
+
"learning_rate": 1.4810606060606061e-05,
|
| 404 |
+
"loss": 0.4343,
|
| 405 |
+
"step": 1175
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 6.82051282051282,
|
| 409 |
+
"grad_norm": 2.47883677482605,
|
| 410 |
+
"learning_rate": 1.4652777777777779e-05,
|
| 411 |
+
"loss": 0.6561,
|
| 412 |
+
"step": 1200
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 6.962962962962963,
|
| 416 |
+
"grad_norm": 52.5629768371582,
|
| 417 |
+
"learning_rate": 1.4494949494949494e-05,
|
| 418 |
+
"loss": 0.6477,
|
| 419 |
+
"step": 1225
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 7.0,
|
| 423 |
+
"eval_accuracy": 0.8490566037735849,
|
| 424 |
+
"eval_loss": 0.6626464128494263,
|
| 425 |
+
"eval_macro_f1": 0.8480730502037321,
|
| 426 |
+
"eval_runtime": 1.4391,
|
| 427 |
+
"eval_samples_per_second": 478.764,
|
| 428 |
+
"eval_steps_per_second": 30.574,
|
| 429 |
+
"eval_weighted_f1": 0.848201027008074,
|
| 430 |
+
"step": 1232
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 7.102564102564102,
|
| 434 |
+
"grad_norm": 9.221817016601562,
|
| 435 |
+
"learning_rate": 1.4337121212121212e-05,
|
| 436 |
+
"loss": 0.3944,
|
| 437 |
+
"step": 1250
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 7.245014245014245,
|
| 441 |
+
"grad_norm": 13.65827465057373,
|
| 442 |
+
"learning_rate": 1.4179292929292931e-05,
|
| 443 |
+
"loss": 0.3707,
|
| 444 |
+
"step": 1275
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 7.387464387464387,
|
| 448 |
+
"grad_norm": 102.12430572509766,
|
| 449 |
+
"learning_rate": 1.4021464646464649e-05,
|
| 450 |
+
"loss": 0.3838,
|
| 451 |
+
"step": 1300
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 7.52991452991453,
|
| 455 |
+
"grad_norm": 41.775047302246094,
|
| 456 |
+
"learning_rate": 1.3863636363636366e-05,
|
| 457 |
+
"loss": 0.3613,
|
| 458 |
+
"step": 1325
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 7.672364672364672,
|
| 462 |
+
"grad_norm": 7.700901985168457,
|
| 463 |
+
"learning_rate": 1.3705808080808084e-05,
|
| 464 |
+
"loss": 0.3836,
|
| 465 |
+
"step": 1350
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 7.814814814814815,
|
| 469 |
+
"grad_norm": 30.64459800720215,
|
| 470 |
+
"learning_rate": 1.35479797979798e-05,
|
| 471 |
+
"loss": 0.4086,
|
| 472 |
+
"step": 1375
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 7.957264957264957,
|
| 476 |
+
"grad_norm": 84.40220642089844,
|
| 477 |
+
"learning_rate": 1.3390151515151517e-05,
|
| 478 |
+
"loss": 0.348,
|
| 479 |
+
"step": 1400
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 8.0,
|
| 483 |
+
"eval_accuracy": 0.8403483309143687,
|
| 484 |
+
"eval_loss": 0.8505121469497681,
|
| 485 |
+
"eval_macro_f1": 0.837636893114618,
|
| 486 |
+
"eval_runtime": 1.5391,
|
| 487 |
+
"eval_samples_per_second": 447.668,
|
| 488 |
+
"eval_steps_per_second": 28.588,
|
| 489 |
+
"eval_weighted_f1": 0.8379419870932212,
|
| 490 |
+
"step": 1408
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 8.096866096866098,
|
| 494 |
+
"grad_norm": 7.839972496032715,
|
| 495 |
+
"learning_rate": 1.3232323232323234e-05,
|
| 496 |
+
"loss": 0.1529,
|
| 497 |
+
"step": 1425
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 8.239316239316238,
|
| 501 |
+
"grad_norm": 147.12326049804688,
|
| 502 |
+
"learning_rate": 1.3074494949494952e-05,
|
| 503 |
+
"loss": 0.2763,
|
| 504 |
+
"step": 1450
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 8.381766381766381,
|
| 508 |
+
"grad_norm": 21.040546417236328,
|
| 509 |
+
"learning_rate": 1.2916666666666668e-05,
|
| 510 |
+
"loss": 0.3735,
|
| 511 |
+
"step": 1475
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 8.524216524216524,
|
| 515 |
+
"grad_norm": 82.0713882446289,
|
| 516 |
+
"learning_rate": 1.2758838383838385e-05,
|
| 517 |
+
"loss": 0.427,
|
| 518 |
+
"step": 1500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 8.666666666666666,
|
| 522 |
+
"grad_norm": 62.74919509887695,
|
| 523 |
+
"learning_rate": 1.2601010101010103e-05,
|
| 524 |
+
"loss": 0.2118,
|
| 525 |
+
"step": 1525
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 8.809116809116809,
|
| 529 |
+
"grad_norm": 0.20923995971679688,
|
| 530 |
+
"learning_rate": 1.244318181818182e-05,
|
| 531 |
+
"loss": 0.5143,
|
| 532 |
+
"step": 1550
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 8.951566951566951,
|
| 536 |
+
"grad_norm": 0.7549470663070679,
|
| 537 |
+
"learning_rate": 1.2285353535353536e-05,
|
| 538 |
+
"loss": 0.4299,
|
| 539 |
+
"step": 1575
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 9.0,
|
| 543 |
+
"eval_accuracy": 0.8258345428156749,
|
| 544 |
+
"eval_loss": 1.0375233888626099,
|
| 545 |
+
"eval_macro_f1": 0.8240438177651709,
|
| 546 |
+
"eval_runtime": 1.4931,
|
| 547 |
+
"eval_samples_per_second": 461.46,
|
| 548 |
+
"eval_steps_per_second": 29.469,
|
| 549 |
+
"eval_weighted_f1": 0.8240742430839567,
|
| 550 |
+
"step": 1584
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 9.091168091168091,
|
| 554 |
+
"grad_norm": 6.6440863609313965,
|
| 555 |
+
"learning_rate": 1.2127525252525254e-05,
|
| 556 |
+
"loss": 0.2515,
|
| 557 |
+
"step": 1600
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 9.233618233618234,
|
| 561 |
+
"grad_norm": 156.8964385986328,
|
| 562 |
+
"learning_rate": 1.1969696969696971e-05,
|
| 563 |
+
"loss": 0.2933,
|
| 564 |
+
"step": 1625
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 9.376068376068377,
|
| 568 |
+
"grad_norm": 10.411484718322754,
|
| 569 |
+
"learning_rate": 1.1811868686868689e-05,
|
| 570 |
+
"loss": 0.3251,
|
| 571 |
+
"step": 1650
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 9.518518518518519,
|
| 575 |
+
"grad_norm": 8.135589599609375,
|
| 576 |
+
"learning_rate": 1.1654040404040406e-05,
|
| 577 |
+
"loss": 0.1041,
|
| 578 |
+
"step": 1675
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 9.660968660968662,
|
| 582 |
+
"grad_norm": 236.67787170410156,
|
| 583 |
+
"learning_rate": 1.1496212121212122e-05,
|
| 584 |
+
"loss": 0.2226,
|
| 585 |
+
"step": 1700
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 9.803418803418804,
|
| 589 |
+
"grad_norm": 97.63804626464844,
|
| 590 |
+
"learning_rate": 1.133838383838384e-05,
|
| 591 |
+
"loss": 0.234,
|
| 592 |
+
"step": 1725
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 9.945868945868947,
|
| 596 |
+
"grad_norm": 0.3027924597263336,
|
| 597 |
+
"learning_rate": 1.1180555555555557e-05,
|
| 598 |
+
"loss": 0.2665,
|
| 599 |
+
"step": 1750
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 10.0,
|
| 603 |
+
"eval_accuracy": 0.8548621190130624,
|
| 604 |
+
"eval_loss": 0.9538707137107849,
|
| 605 |
+
"eval_macro_f1": 0.8546697114483482,
|
| 606 |
+
"eval_runtime": 1.5132,
|
| 607 |
+
"eval_samples_per_second": 455.327,
|
| 608 |
+
"eval_steps_per_second": 29.078,
|
| 609 |
+
"eval_weighted_f1": 0.8548730047689096,
|
| 610 |
+
"step": 1760
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 10.085470085470085,
|
| 614 |
+
"grad_norm": 5.469311237335205,
|
| 615 |
+
"learning_rate": 1.1022727272727275e-05,
|
| 616 |
+
"loss": 0.1478,
|
| 617 |
+
"step": 1775
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 10.227920227920228,
|
| 621 |
+
"grad_norm": 0.15111243724822998,
|
| 622 |
+
"learning_rate": 1.086489898989899e-05,
|
| 623 |
+
"loss": 0.1131,
|
| 624 |
+
"step": 1800
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 10.37037037037037,
|
| 628 |
+
"grad_norm": 14.249789237976074,
|
| 629 |
+
"learning_rate": 1.0707070707070708e-05,
|
| 630 |
+
"loss": 0.1861,
|
| 631 |
+
"step": 1825
|
| 632 |
+
},
|
| 633 |
+
{
|
| 634 |
+
"epoch": 10.512820512820513,
|
| 635 |
+
"grad_norm": 0.0699237734079361,
|
| 636 |
+
"learning_rate": 1.0549242424242425e-05,
|
| 637 |
+
"loss": 0.2356,
|
| 638 |
+
"step": 1850
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 10.655270655270655,
|
| 642 |
+
"grad_norm": 0.058460839092731476,
|
| 643 |
+
"learning_rate": 1.0391414141414143e-05,
|
| 644 |
+
"loss": 0.2,
|
| 645 |
+
"step": 1875
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 10.797720797720798,
|
| 649 |
+
"grad_norm": 0.7808272838592529,
|
| 650 |
+
"learning_rate": 1.0233585858585859e-05,
|
| 651 |
+
"loss": 0.3025,
|
| 652 |
+
"step": 1900
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"epoch": 10.94017094017094,
|
| 656 |
+
"grad_norm": 0.053956158459186554,
|
| 657 |
+
"learning_rate": 1.0075757575757576e-05,
|
| 658 |
+
"loss": 0.0981,
|
| 659 |
+
"step": 1925
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 11.0,
|
| 663 |
+
"eval_accuracy": 0.8403483309143687,
|
| 664 |
+
"eval_loss": 1.0893523693084717,
|
| 665 |
+
"eval_macro_f1": 0.840991946992498,
|
| 666 |
+
"eval_runtime": 1.6394,
|
| 667 |
+
"eval_samples_per_second": 420.282,
|
| 668 |
+
"eval_steps_per_second": 26.839,
|
| 669 |
+
"eval_weighted_f1": 0.8413473160585482,
|
| 670 |
+
"step": 1936
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 11.079772079772079,
|
| 674 |
+
"grad_norm": 0.2754538953304291,
|
| 675 |
+
"learning_rate": 9.917929292929294e-06,
|
| 676 |
+
"loss": 0.1993,
|
| 677 |
+
"step": 1950
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 11.222222222222221,
|
| 681 |
+
"grad_norm": 0.07682927697896957,
|
| 682 |
+
"learning_rate": 9.760101010101011e-06,
|
| 683 |
+
"loss": 0.1484,
|
| 684 |
+
"step": 1975
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 11.364672364672364,
|
| 688 |
+
"grad_norm": 0.044398900121450424,
|
| 689 |
+
"learning_rate": 9.602272727272727e-06,
|
| 690 |
+
"loss": 0.0677,
|
| 691 |
+
"step": 2000
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 11.507122507122507,
|
| 695 |
+
"grad_norm": 0.0840727835893631,
|
| 696 |
+
"learning_rate": 9.444444444444445e-06,
|
| 697 |
+
"loss": 0.2693,
|
| 698 |
+
"step": 2025
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 11.649572649572649,
|
| 702 |
+
"grad_norm": 253.048583984375,
|
| 703 |
+
"learning_rate": 9.286616161616162e-06,
|
| 704 |
+
"loss": 0.214,
|
| 705 |
+
"step": 2050
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 11.792022792022792,
|
| 709 |
+
"grad_norm": 0.06720364838838577,
|
| 710 |
+
"learning_rate": 9.12878787878788e-06,
|
| 711 |
+
"loss": 0.0974,
|
| 712 |
+
"step": 2075
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 11.934472934472934,
|
| 716 |
+
"grad_norm": 0.04520854353904724,
|
| 717 |
+
"learning_rate": 8.970959595959597e-06,
|
| 718 |
+
"loss": 0.1746,
|
| 719 |
+
"step": 2100
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 12.0,
|
| 723 |
+
"eval_accuracy": 0.8447024673439768,
|
| 724 |
+
"eval_loss": 1.077406644821167,
|
| 725 |
+
"eval_macro_f1": 0.844826996232548,
|
| 726 |
+
"eval_runtime": 1.4995,
|
| 727 |
+
"eval_samples_per_second": 459.491,
|
| 728 |
+
"eval_steps_per_second": 29.343,
|
| 729 |
+
"eval_weighted_f1": 0.8452846677856134,
|
| 730 |
+
"step": 2112
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 12.074074074074074,
|
| 734 |
+
"grad_norm": 0.08145175874233246,
|
| 735 |
+
"learning_rate": 8.813131313131313e-06,
|
| 736 |
+
"loss": 0.1545,
|
| 737 |
+
"step": 2125
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 12.216524216524217,
|
| 741 |
+
"grad_norm": 3.1033504009246826,
|
| 742 |
+
"learning_rate": 8.65530303030303e-06,
|
| 743 |
+
"loss": 0.1476,
|
| 744 |
+
"step": 2150
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 12.35897435897436,
|
| 748 |
+
"grad_norm": 0.06478939205408096,
|
| 749 |
+
"learning_rate": 8.497474747474748e-06,
|
| 750 |
+
"loss": 0.0822,
|
| 751 |
+
"step": 2175
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 12.501424501424502,
|
| 755 |
+
"grad_norm": 0.11002465337514877,
|
| 756 |
+
"learning_rate": 8.339646464646465e-06,
|
| 757 |
+
"loss": 0.1695,
|
| 758 |
+
"step": 2200
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 12.643874643874645,
|
| 762 |
+
"grad_norm": 16.418249130249023,
|
| 763 |
+
"learning_rate": 8.181818181818183e-06,
|
| 764 |
+
"loss": 0.2139,
|
| 765 |
+
"step": 2225
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 12.786324786324787,
|
| 769 |
+
"grad_norm": 0.05900779739022255,
|
| 770 |
+
"learning_rate": 8.0239898989899e-06,
|
| 771 |
+
"loss": 0.1104,
|
| 772 |
+
"step": 2250
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 12.928774928774928,
|
| 776 |
+
"grad_norm": 271.0628967285156,
|
| 777 |
+
"learning_rate": 7.866161616161618e-06,
|
| 778 |
+
"loss": 0.1361,
|
| 779 |
+
"step": 2275
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 13.0,
|
| 783 |
+
"eval_accuracy": 0.8447024673439768,
|
| 784 |
+
"eval_loss": 1.0756146907806396,
|
| 785 |
+
"eval_macro_f1": 0.8442854007719814,
|
| 786 |
+
"eval_runtime": 1.2397,
|
| 787 |
+
"eval_samples_per_second": 555.773,
|
| 788 |
+
"eval_steps_per_second": 35.492,
|
| 789 |
+
"eval_weighted_f1": 0.8447286563535076,
|
| 790 |
+
"step": 2288
|
| 791 |
+
}
|
| 792 |
+
],
|
| 793 |
+
"logging_steps": 25,
|
| 794 |
+
"max_steps": 3520,
|
| 795 |
+
"num_input_tokens_seen": 0,
|
| 796 |
+
"num_train_epochs": 20,
|
| 797 |
+
"save_steps": 500,
|
| 798 |
+
"stateful_callbacks": {
|
| 799 |
+
"EarlyStoppingCallback": {
|
| 800 |
+
"args": {
|
| 801 |
+
"early_stopping_patience": 3,
|
| 802 |
+
"early_stopping_threshold": 0.0
|
| 803 |
+
},
|
| 804 |
+
"attributes": {
|
| 805 |
+
"early_stopping_patience_counter": 3
|
| 806 |
+
}
|
| 807 |
+
},
|
| 808 |
+
"TrainerControl": {
|
| 809 |
+
"args": {
|
| 810 |
+
"should_epoch_stop": false,
|
| 811 |
+
"should_evaluate": false,
|
| 812 |
+
"should_log": false,
|
| 813 |
+
"should_save": true,
|
| 814 |
+
"should_training_stop": true
|
| 815 |
+
},
|
| 816 |
+
"attributes": {}
|
| 817 |
+
}
|
| 818 |
+
},
|
| 819 |
+
"total_flos": 0.0,
|
| 820 |
+
"train_batch_size": 8,
|
| 821 |
+
"trial_name": null,
|
| 822 |
+
"trial_params": null
|
| 823 |
+
}
|
ood-vifactcheck-bamibert-source-source-04-s43-20260829T011110Z-3858bb98/trainer/checkpoint-1408/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-source-source-04-s43-20260829T011110Z-3858bb98/trainer/checkpoint-1408/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-vifactcheck-bamibert-source-source-04-s43-20260829T011110Z-3858bb98/trainer/checkpoint-1408/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-source-source-04-s43-20260829T011110Z-3858bb98/trainer/checkpoint-1408/trainer_state.json
ADDED
|
@@ -0,0 +1,523 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 880,
|
| 3 |
+
"best_metric": 0.8445520325632899,
|
| 4 |
+
"best_model_checkpoint": "/root/VNEncoder/results/checkpoints/ood-vifactcheck-bamibert-source-source-04-s43-20260829T011110Z-3858bb98/trainer/checkpoint-880",
|
| 5 |
+
"epoch": 8.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1408,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.14245014245014245,
|
| 14 |
+
"grad_norm": 27.754772186279297,
|
| 15 |
+
"learning_rate": 1.3636363636363636e-06,
|
| 16 |
+
"loss": 4.4025,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.2849002849002849,
|
| 21 |
+
"grad_norm": 14.381654739379883,
|
| 22 |
+
"learning_rate": 2.784090909090909e-06,
|
| 23 |
+
"loss": 4.4236,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.42735042735042733,
|
| 28 |
+
"grad_norm": 14.20284366607666,
|
| 29 |
+
"learning_rate": 4.204545454545455e-06,
|
| 30 |
+
"loss": 4.3777,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.5698005698005698,
|
| 35 |
+
"grad_norm": 11.914273262023926,
|
| 36 |
+
"learning_rate": 5.625e-06,
|
| 37 |
+
"loss": 4.4111,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.7122507122507122,
|
| 42 |
+
"grad_norm": 10.217042922973633,
|
| 43 |
+
"learning_rate": 7.045454545454546e-06,
|
| 44 |
+
"loss": 4.3255,
|
| 45 |
+
"step": 125
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.8547008547008547,
|
| 49 |
+
"grad_norm": 10.928240776062012,
|
| 50 |
+
"learning_rate": 8.465909090909091e-06,
|
| 51 |
+
"loss": 4.3885,
|
| 52 |
+
"step": 150
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.9971509971509972,
|
| 56 |
+
"grad_norm": 14.12752914428711,
|
| 57 |
+
"learning_rate": 9.886363636363637e-06,
|
| 58 |
+
"loss": 4.3139,
|
| 59 |
+
"step": 175
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 1.0,
|
| 63 |
+
"eval_accuracy": 0.39622641509433965,
|
| 64 |
+
"eval_loss": 1.063523292541504,
|
| 65 |
+
"eval_macro_f1": 0.27879523865511663,
|
| 66 |
+
"eval_runtime": 1.169,
|
| 67 |
+
"eval_samples_per_second": 589.392,
|
| 68 |
+
"eval_steps_per_second": 37.639,
|
| 69 |
+
"eval_weighted_f1": 0.2858781262303967,
|
| 70 |
+
"step": 176
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 1.1367521367521367,
|
| 74 |
+
"grad_norm": 18.73723602294922,
|
| 75 |
+
"learning_rate": 1.1306818181818183e-05,
|
| 76 |
+
"loss": 4.1503,
|
| 77 |
+
"step": 200
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.2792022792022792,
|
| 81 |
+
"grad_norm": 23.313190460205078,
|
| 82 |
+
"learning_rate": 1.2727272727272728e-05,
|
| 83 |
+
"loss": 4.0998,
|
| 84 |
+
"step": 225
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.4216524216524218,
|
| 88 |
+
"grad_norm": 24.607540130615234,
|
| 89 |
+
"learning_rate": 1.4147727272727274e-05,
|
| 90 |
+
"loss": 3.6018,
|
| 91 |
+
"step": 250
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.564102564102564,
|
| 95 |
+
"grad_norm": 46.314247131347656,
|
| 96 |
+
"learning_rate": 1.5568181818181822e-05,
|
| 97 |
+
"loss": 3.2531,
|
| 98 |
+
"step": 275
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 1.7065527065527064,
|
| 102 |
+
"grad_norm": 26.79118537902832,
|
| 103 |
+
"learning_rate": 1.6988636363636368e-05,
|
| 104 |
+
"loss": 2.7455,
|
| 105 |
+
"step": 300
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 1.849002849002849,
|
| 109 |
+
"grad_norm": 47.866783142089844,
|
| 110 |
+
"learning_rate": 1.840909090909091e-05,
|
| 111 |
+
"loss": 2.7255,
|
| 112 |
+
"step": 325
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 1.9914529914529915,
|
| 116 |
+
"grad_norm": 35.993404388427734,
|
| 117 |
+
"learning_rate": 1.9829545454545456e-05,
|
| 118 |
+
"loss": 2.6046,
|
| 119 |
+
"step": 350
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 2.0,
|
| 123 |
+
"eval_accuracy": 0.7982583454281568,
|
| 124 |
+
"eval_loss": 0.5425768494606018,
|
| 125 |
+
"eval_macro_f1": 0.7962094386718523,
|
| 126 |
+
"eval_runtime": 1.1477,
|
| 127 |
+
"eval_samples_per_second": 600.33,
|
| 128 |
+
"eval_steps_per_second": 38.338,
|
| 129 |
+
"eval_weighted_f1": 0.7968294277311166,
|
| 130 |
+
"step": 352
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 2.131054131054131,
|
| 134 |
+
"grad_norm": 51.88352966308594,
|
| 135 |
+
"learning_rate": 1.9861111111111114e-05,
|
| 136 |
+
"loss": 2.0865,
|
| 137 |
+
"step": 375
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 2.2735042735042734,
|
| 141 |
+
"grad_norm": 35.94597244262695,
|
| 142 |
+
"learning_rate": 1.9703282828282832e-05,
|
| 143 |
+
"loss": 2.1075,
|
| 144 |
+
"step": 400
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 2.415954415954416,
|
| 148 |
+
"grad_norm": 34.07805252075195,
|
| 149 |
+
"learning_rate": 1.9545454545454546e-05,
|
| 150 |
+
"loss": 2.1175,
|
| 151 |
+
"step": 425
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 2.5584045584045585,
|
| 155 |
+
"grad_norm": 72.2060775756836,
|
| 156 |
+
"learning_rate": 1.9387626262626263e-05,
|
| 157 |
+
"loss": 1.8655,
|
| 158 |
+
"step": 450
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.700854700854701,
|
| 162 |
+
"grad_norm": 58.4161376953125,
|
| 163 |
+
"learning_rate": 1.922979797979798e-05,
|
| 164 |
+
"loss": 1.7381,
|
| 165 |
+
"step": 475
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 2.8433048433048436,
|
| 169 |
+
"grad_norm": 46.373355865478516,
|
| 170 |
+
"learning_rate": 1.90719696969697e-05,
|
| 171 |
+
"loss": 1.9878,
|
| 172 |
+
"step": 500
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 2.9857549857549857,
|
| 176 |
+
"grad_norm": 25.535076141357422,
|
| 177 |
+
"learning_rate": 1.8914141414141416e-05,
|
| 178 |
+
"loss": 1.9817,
|
| 179 |
+
"step": 525
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 3.0,
|
| 183 |
+
"eval_accuracy": 0.841799709724238,
|
| 184 |
+
"eval_loss": 0.4397426247596741,
|
| 185 |
+
"eval_macro_f1": 0.841679897973509,
|
| 186 |
+
"eval_runtime": 1.1585,
|
| 187 |
+
"eval_samples_per_second": 594.754,
|
| 188 |
+
"eval_steps_per_second": 37.981,
|
| 189 |
+
"eval_weighted_f1": 0.8418365441025588,
|
| 190 |
+
"step": 528
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 3.1253561253561255,
|
| 194 |
+
"grad_norm": 68.33673095703125,
|
| 195 |
+
"learning_rate": 1.8756313131313133e-05,
|
| 196 |
+
"loss": 1.2574,
|
| 197 |
+
"step": 550
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 3.267806267806268,
|
| 201 |
+
"grad_norm": 17.43631362915039,
|
| 202 |
+
"learning_rate": 1.859848484848485e-05,
|
| 203 |
+
"loss": 1.3244,
|
| 204 |
+
"step": 575
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 3.41025641025641,
|
| 208 |
+
"grad_norm": 54.527000427246094,
|
| 209 |
+
"learning_rate": 1.844065656565657e-05,
|
| 210 |
+
"loss": 1.7422,
|
| 211 |
+
"step": 600
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 3.5527065527065527,
|
| 215 |
+
"grad_norm": 28.50528907775879,
|
| 216 |
+
"learning_rate": 1.8282828282828286e-05,
|
| 217 |
+
"loss": 1.595,
|
| 218 |
+
"step": 625
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.695156695156695,
|
| 222 |
+
"grad_norm": 57.27180480957031,
|
| 223 |
+
"learning_rate": 1.8125e-05,
|
| 224 |
+
"loss": 1.5379,
|
| 225 |
+
"step": 650
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.8376068376068377,
|
| 229 |
+
"grad_norm": 76.50605010986328,
|
| 230 |
+
"learning_rate": 1.7967171717171718e-05,
|
| 231 |
+
"loss": 1.2609,
|
| 232 |
+
"step": 675
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 3.98005698005698,
|
| 236 |
+
"grad_norm": 40.559295654296875,
|
| 237 |
+
"learning_rate": 1.7809343434343435e-05,
|
| 238 |
+
"loss": 1.4838,
|
| 239 |
+
"step": 700
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 4.0,
|
| 243 |
+
"eval_accuracy": 0.8403483309143687,
|
| 244 |
+
"eval_loss": 0.4359947741031647,
|
| 245 |
+
"eval_macro_f1": 0.8405721842844515,
|
| 246 |
+
"eval_runtime": 1.1748,
|
| 247 |
+
"eval_samples_per_second": 586.502,
|
| 248 |
+
"eval_steps_per_second": 37.454,
|
| 249 |
+
"eval_weighted_f1": 0.8410506840735219,
|
| 250 |
+
"step": 704
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 4.119658119658119,
|
| 254 |
+
"grad_norm": 24.998632431030273,
|
| 255 |
+
"learning_rate": 1.7651515151515153e-05,
|
| 256 |
+
"loss": 0.9559,
|
| 257 |
+
"step": 725
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 4.262108262108262,
|
| 261 |
+
"grad_norm": 30.147058486938477,
|
| 262 |
+
"learning_rate": 1.749368686868687e-05,
|
| 263 |
+
"loss": 0.8775,
|
| 264 |
+
"step": 750
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 4.404558404558404,
|
| 268 |
+
"grad_norm": 72.00198364257812,
|
| 269 |
+
"learning_rate": 1.7335858585858588e-05,
|
| 270 |
+
"loss": 0.9006,
|
| 271 |
+
"step": 775
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 4.547008547008547,
|
| 275 |
+
"grad_norm": 44.2113151550293,
|
| 276 |
+
"learning_rate": 1.7178030303030305e-05,
|
| 277 |
+
"loss": 0.9198,
|
| 278 |
+
"step": 800
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 4.689458689458689,
|
| 282 |
+
"grad_norm": 23.130006790161133,
|
| 283 |
+
"learning_rate": 1.7020202020202023e-05,
|
| 284 |
+
"loss": 0.7771,
|
| 285 |
+
"step": 825
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 4.831908831908832,
|
| 289 |
+
"grad_norm": 46.27170181274414,
|
| 290 |
+
"learning_rate": 1.6862373737373737e-05,
|
| 291 |
+
"loss": 0.8763,
|
| 292 |
+
"step": 850
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 4.9743589743589745,
|
| 296 |
+
"grad_norm": 85.02490234375,
|
| 297 |
+
"learning_rate": 1.6704545454545454e-05,
|
| 298 |
+
"loss": 0.8004,
|
| 299 |
+
"step": 875
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 5.0,
|
| 303 |
+
"eval_accuracy": 0.8447024673439768,
|
| 304 |
+
"eval_loss": 0.4922022521495819,
|
| 305 |
+
"eval_macro_f1": 0.8445520325632899,
|
| 306 |
+
"eval_runtime": 1.2679,
|
| 307 |
+
"eval_samples_per_second": 543.421,
|
| 308 |
+
"eval_steps_per_second": 34.703,
|
| 309 |
+
"eval_weighted_f1": 0.8453273906740522,
|
| 310 |
+
"step": 880
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 5.113960113960114,
|
| 314 |
+
"grad_norm": 56.17656326293945,
|
| 315 |
+
"learning_rate": 1.6546717171717172e-05,
|
| 316 |
+
"loss": 0.7194,
|
| 317 |
+
"step": 900
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 5.256410256410256,
|
| 321 |
+
"grad_norm": 65.36344146728516,
|
| 322 |
+
"learning_rate": 1.638888888888889e-05,
|
| 323 |
+
"loss": 0.6272,
|
| 324 |
+
"step": 925
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 5.398860398860399,
|
| 328 |
+
"grad_norm": 57.312896728515625,
|
| 329 |
+
"learning_rate": 1.6231060606060607e-05,
|
| 330 |
+
"loss": 0.6486,
|
| 331 |
+
"step": 950
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.5413105413105415,
|
| 335 |
+
"grad_norm": 12.53235149383545,
|
| 336 |
+
"learning_rate": 1.6073232323232324e-05,
|
| 337 |
+
"loss": 0.5263,
|
| 338 |
+
"step": 975
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 5.683760683760684,
|
| 342 |
+
"grad_norm": 52.88577651977539,
|
| 343 |
+
"learning_rate": 1.5915404040404042e-05,
|
| 344 |
+
"loss": 0.6936,
|
| 345 |
+
"step": 1000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 5.8262108262108265,
|
| 349 |
+
"grad_norm": 71.69739532470703,
|
| 350 |
+
"learning_rate": 1.575757575757576e-05,
|
| 351 |
+
"loss": 0.7758,
|
| 352 |
+
"step": 1025
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 5.968660968660968,
|
| 356 |
+
"grad_norm": 57.650535583496094,
|
| 357 |
+
"learning_rate": 1.5599747474747477e-05,
|
| 358 |
+
"loss": 0.4996,
|
| 359 |
+
"step": 1050
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 6.0,
|
| 363 |
+
"eval_accuracy": 0.8171262699564587,
|
| 364 |
+
"eval_loss": 0.687231183052063,
|
| 365 |
+
"eval_macro_f1": 0.8195741704437357,
|
| 366 |
+
"eval_runtime": 1.1904,
|
| 367 |
+
"eval_samples_per_second": 578.783,
|
| 368 |
+
"eval_steps_per_second": 36.961,
|
| 369 |
+
"eval_weighted_f1": 0.8205242871444677,
|
| 370 |
+
"step": 1056
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 6.1082621082621085,
|
| 374 |
+
"grad_norm": 3.100119113922119,
|
| 375 |
+
"learning_rate": 1.544191919191919e-05,
|
| 376 |
+
"loss": 0.4165,
|
| 377 |
+
"step": 1075
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 6.250712250712251,
|
| 381 |
+
"grad_norm": 119.09585571289062,
|
| 382 |
+
"learning_rate": 1.528409090909091e-05,
|
| 383 |
+
"loss": 0.5148,
|
| 384 |
+
"step": 1100
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 6.3931623931623935,
|
| 388 |
+
"grad_norm": 47.853736877441406,
|
| 389 |
+
"learning_rate": 1.5126262626262626e-05,
|
| 390 |
+
"loss": 0.6765,
|
| 391 |
+
"step": 1125
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 6.535612535612536,
|
| 395 |
+
"grad_norm": 104.02178192138672,
|
| 396 |
+
"learning_rate": 1.4968434343434344e-05,
|
| 397 |
+
"loss": 0.5269,
|
| 398 |
+
"step": 1150
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 6.678062678062678,
|
| 402 |
+
"grad_norm": 41.17183303833008,
|
| 403 |
+
"learning_rate": 1.4810606060606061e-05,
|
| 404 |
+
"loss": 0.7421,
|
| 405 |
+
"step": 1175
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 6.82051282051282,
|
| 409 |
+
"grad_norm": 43.6857795715332,
|
| 410 |
+
"learning_rate": 1.4652777777777779e-05,
|
| 411 |
+
"loss": 0.4717,
|
| 412 |
+
"step": 1200
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 6.962962962962963,
|
| 416 |
+
"grad_norm": 39.86498260498047,
|
| 417 |
+
"learning_rate": 1.4494949494949494e-05,
|
| 418 |
+
"loss": 0.5544,
|
| 419 |
+
"step": 1225
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 7.0,
|
| 423 |
+
"eval_accuracy": 0.8301886792452831,
|
| 424 |
+
"eval_loss": 0.8011407256126404,
|
| 425 |
+
"eval_macro_f1": 0.8301066526878245,
|
| 426 |
+
"eval_runtime": 1.1477,
|
| 427 |
+
"eval_samples_per_second": 600.322,
|
| 428 |
+
"eval_steps_per_second": 38.337,
|
| 429 |
+
"eval_weighted_f1": 0.8297965932874872,
|
| 430 |
+
"step": 1232
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 7.102564102564102,
|
| 434 |
+
"grad_norm": 100.96305847167969,
|
| 435 |
+
"learning_rate": 1.4337121212121212e-05,
|
| 436 |
+
"loss": 0.2924,
|
| 437 |
+
"step": 1250
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 7.245014245014245,
|
| 441 |
+
"grad_norm": 5.0722856521606445,
|
| 442 |
+
"learning_rate": 1.4179292929292931e-05,
|
| 443 |
+
"loss": 0.2871,
|
| 444 |
+
"step": 1275
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 7.387464387464387,
|
| 448 |
+
"grad_norm": 106.19259643554688,
|
| 449 |
+
"learning_rate": 1.4021464646464649e-05,
|
| 450 |
+
"loss": 0.3225,
|
| 451 |
+
"step": 1300
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 7.52991452991453,
|
| 455 |
+
"grad_norm": 0.735567569732666,
|
| 456 |
+
"learning_rate": 1.3863636363636366e-05,
|
| 457 |
+
"loss": 0.3055,
|
| 458 |
+
"step": 1325
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 7.672364672364672,
|
| 462 |
+
"grad_norm": 92.68929290771484,
|
| 463 |
+
"learning_rate": 1.3705808080808084e-05,
|
| 464 |
+
"loss": 0.296,
|
| 465 |
+
"step": 1350
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 7.814814814814815,
|
| 469 |
+
"grad_norm": 117.81938934326172,
|
| 470 |
+
"learning_rate": 1.35479797979798e-05,
|
| 471 |
+
"loss": 0.7087,
|
| 472 |
+
"step": 1375
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 7.957264957264957,
|
| 476 |
+
"grad_norm": 0.6199210286140442,
|
| 477 |
+
"learning_rate": 1.3390151515151517e-05,
|
| 478 |
+
"loss": 0.4125,
|
| 479 |
+
"step": 1400
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 8.0,
|
| 483 |
+
"eval_accuracy": 0.8403483309143687,
|
| 484 |
+
"eval_loss": 0.9340547323226929,
|
| 485 |
+
"eval_macro_f1": 0.8401992086745299,
|
| 486 |
+
"eval_runtime": 1.1668,
|
| 487 |
+
"eval_samples_per_second": 590.495,
|
| 488 |
+
"eval_steps_per_second": 37.709,
|
| 489 |
+
"eval_weighted_f1": 0.8403873603670763,
|
| 490 |
+
"step": 1408
|
| 491 |
+
}
|
| 492 |
+
],
|
| 493 |
+
"logging_steps": 25,
|
| 494 |
+
"max_steps": 3520,
|
| 495 |
+
"num_input_tokens_seen": 0,
|
| 496 |
+
"num_train_epochs": 20,
|
| 497 |
+
"save_steps": 500,
|
| 498 |
+
"stateful_callbacks": {
|
| 499 |
+
"EarlyStoppingCallback": {
|
| 500 |
+
"args": {
|
| 501 |
+
"early_stopping_patience": 3,
|
| 502 |
+
"early_stopping_threshold": 0.0
|
| 503 |
+
},
|
| 504 |
+
"attributes": {
|
| 505 |
+
"early_stopping_patience_counter": 3
|
| 506 |
+
}
|
| 507 |
+
},
|
| 508 |
+
"TrainerControl": {
|
| 509 |
+
"args": {
|
| 510 |
+
"should_epoch_stop": false,
|
| 511 |
+
"should_evaluate": false,
|
| 512 |
+
"should_log": false,
|
| 513 |
+
"should_save": true,
|
| 514 |
+
"should_training_stop": true
|
| 515 |
+
},
|
| 516 |
+
"attributes": {}
|
| 517 |
+
}
|
| 518 |
+
},
|
| 519 |
+
"total_flos": 0.0,
|
| 520 |
+
"train_batch_size": 8,
|
| 521 |
+
"trial_name": null,
|
| 522 |
+
"trial_params": null
|
| 523 |
+
}
|
ood-vifactcheck-bamibert-source-source-04-s43-20260829T011110Z-3858bb98/trainer/checkpoint-880/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-source-source-04-s43-20260829T011110Z-3858bb98/trainer/checkpoint-880/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
ood-vifactcheck-bamibert-source-source-04-s43-20260829T011110Z-3858bb98/trainer/checkpoint-880/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 |
+
}
|