2025-07-10 11:07:44,783 - INFO - machamp.model.trainer - cmd: train.py --dataset_configs configs/ota_config.json --device 0
2025-07-10 11:07:44,823 - INFO - machamp.model.trainer - GPU: Tesla T4
2025-07-10 11:07:44,824 - INFO - machamp.model.trainer - Torch version: 2.6.0+cu124
2025-07-10 11:07:44,824 - INFO - machamp.model.trainer - Transformers version: 4.53.1
2025-07-10 11:07:44,824 - INFO - machamp.model.trainer - MaChAmp git version 75fbbee9db00154bfefb3b0e8e1f5e1dae60694f
2025-07-10 11:07:45,026 - ERROR - STDERR - 
2025-07-10 11:07:45,026 - ERROR - STDERR - tokenizer_config.json: 0.00B [00:00, ?B/s]
2025-07-10 11:07:45,027 - ERROR - STDERR - 
2025-07-10 11:07:45,027 - ERROR - STDERR - tokenizer_config.json: 1.18kB [00:00, 6.50MB/s]
2025-07-10 11:07:45,385 - ERROR - STDERR - 
2025-07-10 11:07:45,386 - ERROR - STDERR - sentencepiece.bpe.model:   0%|          | 0.00/5.07M [00:00<?, ?B/s]
2025-07-10 11:07:45,489 - ERROR - STDERR - 
2025-07-10 11:07:45,489 - ERROR - STDERR - sentencepiece.bpe.model: 100%|##########| 5.07M/5.07M [00:00<00:00, 49.2MB/s]
2025-07-10 11:07:45,492 - ERROR - STDERR - 
2025-07-10 11:07:45,492 - ERROR - STDERR - sentencepiece.bpe.model: 100%|##########| 5.07M/5.07M [00:00<00:00, 47.8MB/s]
2025-07-10 11:07:45,751 - ERROR - STDERR - 
2025-07-10 11:07:45,751 - ERROR - STDERR - special_tokens_map.json:   0%|          | 0.00/280 [00:00<?, ?B/s]
2025-07-10 11:07:45,751 - ERROR - STDERR - 
2025-07-10 11:07:45,751 - ERROR - STDERR - special_tokens_map.json: 100%|##########| 280/280 [00:00<00:00, 2.81MB/s]
2025-07-10 11:07:46,029 - ERROR - STDERR - 
2025-07-10 11:07:46,029 - ERROR - STDERR - tokenizer.json:   0%|          | 0.00/17.1M [00:00<?, ?B/s]
2025-07-10 11:07:46,278 - ERROR - STDERR - 
2025-07-10 11:07:46,278 - ERROR - STDERR - tokenizer.json: 100%|##########| 17.1M/17.1M [00:00<00:00, 68.9MB/s]
2025-07-10 11:07:46,282 - ERROR - STDERR - 
2025-07-10 11:07:46,282 - ERROR - STDERR - tokenizer.json: 100%|##########| 17.1M/17.1M [00:00<00:00, 67.7MB/s]
2025-07-10 11:07:47,568 - INFO - machamp.data.machamp_dataset - Reading /content/train.conllu...
2025-07-10 11:07:51,783 - INFO - machamp.readers.read_sequence - Stats UD-EWT (/content/train.conllu):
2025-07-10 11:07:51,786 - INFO - machamp.readers.read_sequence - Lines:      1,164
2025-07-10 11:07:51,786 - INFO - machamp.readers.read_sequence - Words:      11,351
2025-07-10 11:07:51,786 - INFO - machamp.readers.read_sequence - Subwords:   34,185
2025-07-10 11:07:51,786 - INFO - machamp.readers.read_sequence - Unks:       563
2025-07-10 11:07:51,786 - INFO - machamp.readers.read_sequence - Pre-splits: 0
2025-07-10 11:07:51,787 - INFO - machamp.data.machamp_dataset - Done reading /content/train.conllu (4.0s)

2025-07-10 11:07:53,229 - INFO - machamp.data.machamp_dataset - Reading /content/dev.conllu...
2025-07-10 11:07:53,244 - WARNING - machamp.data.machamp_vocabulary - Case=Acc|Number=Sing|Number[psor]=Sing|Person=3|Person[psor]=3|Polarity=Pos can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,284 - WARNING - machamp.data.machamp_vocabulary - Mood=Opt|Number=Plur|Person=1|Polarity=Neg|Tense=Pres|VerbForm=Fin can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,319 - WARNING - machamp.data.machamp_vocabulary - Aspect=Perf|Evident=Fh|Mood=Ind|Number=Plur|Person=3|Polarity=Pos|Tense=Past|VerbForm=Fin|Voice=Cau can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,351 - WARNING - machamp.data.machamp_vocabulary - Evident=Nfh|Number=Sing|Person=3|Polarity=Pos|Tense=Past|VerbForm=Fin can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,378 - WARNING - machamp.data.machamp_vocabulary - Aspect=Perf|Evident=Fh|Mood=Cnd|Number=Sing|Person=2|Polarity=Neg|Tense=Past|VerbForm=Fin can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,380 - WARNING - machamp.data.machamp_vocabulary - Mood=Ind|Number=Sing|Person=3|Tense=Pres|VerbForm=Fin can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,406 - WARNING - machamp.data.machamp_vocabulary - Aspect=Perf|Case=Nom|Number[psor]=Sing|Person[psor]=3|Polarity=Pos|Tense=Past|VerbForm=Part can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,460 - WARNING - machamp.data.machamp_vocabulary - Case=Nom|Number=Plur|Number[psor]=Plur|Person=3|Person[psor]=3|PronType=Prs|Reflex=Yes can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,476 - WARNING - machamp.data.machamp_vocabulary - Aspect=Imp|Polarity=Pos|Tense=Fut|VerbForm=Part can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,482 - WARNING - machamp.data.machamp_vocabulary - Case=Acc|Number=Sing|Number[psor]=Plur|Person=3|Person[psor]=3|Polarity=Pos|VerbForm=Vnoun can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,493 - WARNING - machamp.data.machamp_vocabulary - Case=Loc|Gender=Masc|Number=Sing|Person=3 can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,493 - WARNING - machamp.data.machamp_vocabulary - Aspect=Hab|Polarity=Neg|Tense=Pres|VerbForm=Part can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,500 - WARNING - machamp.data.machamp_vocabulary - Case=Nom|Number=Plur|Person=2 can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,511 - WARNING - machamp.data.machamp_vocabulary - Evident=Nfh|Number=Sing|Person=3|Polarity=Pos|Tense=Past|VerbForm=Fin can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,552 - WARNING - machamp.data.machamp_vocabulary - Case=Ins|Gender=Fem|Number=Sing|Number[psor]=Sing|Person=3|Person[psor]=3 can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,552 - WARNING - machamp.data.machamp_vocabulary - Aspect=Perf|Mood=Ind|Number=Plur|Person=3|Polarity=Pos|Tense=Past|VerbForm=Fin can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,556 - WARNING - machamp.data.machamp_vocabulary - Aspect=Perf|Mood=Ind|Number=Sing|Person=3|Polarity=Pos|Tense=Past|VerbForm=Fin|Voice=Pass can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,560 - WARNING - machamp.data.machamp_vocabulary - Aspect=Hab|Polarity=Neg|Tense=Pres|VerbForm=Part can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,584 - WARNING - machamp.data.machamp_vocabulary - Aspect=Perf|Case=Nom|Number[psor]=Sing|Person[psor]=3|Polarity=Pos|Tense=Pres|VerbForm=Vnoun can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,602 - WARNING - machamp.data.machamp_vocabulary - Case=Nom|Number=Plur|Person=3|Typo=Yes can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,641 - WARNING - machamp.data.machamp_vocabulary - Case=Acc|Gender=Masc|Number=Sing|Number[psor]=Sing|Person=3|Person[psor]=3|PronType=Ind can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,649 - WARNING - machamp.data.machamp_vocabulary - Without can not be found in namespace xpos. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,662 - WARNING - machamp.data.machamp_vocabulary - Aspect=Perf|Number[psor]=Sing|Person[psor]=3|Polarity=Pos|Tense=Past|VerbForm=Part|Voice=Pass can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,670 - WARNING - machamp.data.machamp_vocabulary - Case=Nom|Number=Sing|Number[psor]=Plur|Person=3|Person[psor]=3|Polarity=Pos|Tense=Pres|VerbForm=Part can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,683 - WARNING - machamp.data.machamp_vocabulary - Aspect=Hab|Polarity=Neg|VerbForm=Part can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,683 - WARNING - machamp.data.machamp_vocabulary - Aspect=Perf|Evident=Fh|Mood=Opt|Number=Sing|Person=3|Polarity=Pos|Tense=Past|VerbForm=Fin can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,706 - WARNING - machamp.data.machamp_vocabulary - Case=Loc|Number=Sing|Person=3|Polarity=Neg|VerbForm=Vnoun can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,711 - WARNING - machamp.data.machamp_vocabulary - Aspect=Hab|Mood=Cnd|Number=Sing|Person=3|Polarity=Neg|Tense=Pres|VerbForm=Fin can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,722 - WARNING - machamp.data.machamp_vocabulary - Case=Abl|Number=Sing|Person=3|Polarity=Pos|Tense=Pres|VerbForm=Part can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,733 - WARNING - machamp.data.machamp_vocabulary - Evident=Nfh|Number=Sing|Person=3|Polarity=Neg|Tense=Past|VerbForm=Fin can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,743 - WARNING - machamp.data.machamp_vocabulary - Case=Gen|Number=Sing|Number[psor]=Sing|Person=3|Person[psor]=3|PronType=Dem can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,750 - WARNING - machamp.data.machamp_vocabulary - Evident=Nfh|Number=Sing|Person=3|Polarity=Neg|Tense=Past|VerbForm=Part can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,753 - WARNING - machamp.data.machamp_vocabulary - Case=Abl|Number=Sing|Person=3|PronType=Ind can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,832 - WARNING - machamp.data.machamp_vocabulary - Case=Acc|Number=Sing|Number[psor]=Sing|Person=3|Person[psor]=3|Polarity=Pos|VerbForm=Vnoun|Voice=Pass can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,855 - WARNING - machamp.data.machamp_vocabulary - Aspect=Perf|Evident=Fh|Number=Sing|Person=3|Polarity=Pos|Tense=Past|VerbForm=Fin can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,874 - WARNING - machamp.data.machamp_vocabulary - Case=Nom|Gender=Fem|Number=Dual|Person=3 can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,874 - WARNING - machamp.data.machamp_vocabulary - Gender=Fem|Number=Dual can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,891 - WARNING - machamp.data.machamp_vocabulary - Aspect=Hab|Mood=Ind|Number=Plur|Person=3|Polarity=Pos|Tense=Pres|VerbForm=Fin|Voice=Pass can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,922 - WARNING - machamp.data.machamp_vocabulary - Case=Acc|Number=Sing|Number[psor]=Sing|Person=3|Person[psor]=2|Polarity=Pos|VerbForm=Vnoun can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,925 - WARNING - machamp.data.machamp_vocabulary - Evident=Nfh|Mood=Ind|Number=Plur|Person=3|Polarity=Pos|Tense=Past|VerbForm=Fin can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:53,992 - WARNING - machamp.data.machamp_vocabulary - Aspect=Hab|Mood=Cnd|Number=Sing|Person=1|Polarity=Neg|Tense=Pres|VerbForm=Fin can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:54,005 - WARNING - machamp.data.machamp_vocabulary - Aspect=Perf|Case=Dat|Number=Plur|Number[psor]=Plur|Person=3|Person[psor]=3|Polarity=Neg|Tense=Past|VerbForm=Part can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:54,032 - WARNING - machamp.data.machamp_vocabulary - Number=Sing|Person=2|Polarity=Pos|Tense=Pres can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:54,093 - WARNING - machamp.data.machamp_vocabulary - Aspect=Prosp|Mood=Ind|Polarity=Pos|Tense=Fut|VerbForm=Part can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:54,106 - WARNING - machamp.data.machamp_vocabulary - Case=Ins|Gender=Fem|Number=Sing|Number[psor]=Sing|Person=3|Person[psor]=3 can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:54,114 - WARNING - machamp.data.machamp_vocabulary - Aspect=Imp|Mood=Ind|Number=Sing|Person=3|Polarity=Pos|VerbForm=Fin can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:54,144 - WARNING - machamp.data.machamp_vocabulary - Case=Ins|Number=Sing|Person=3|Polarity=Neg|VerbForm=Vnoun can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:54,171 - WARNING - machamp.data.machamp_vocabulary - Case=Ins|Gender=Fem|Number=Sing|Number[psor]=Sing|Person=3|Person[psor]=1 can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:54,171 - WARNING - machamp.data.machamp_vocabulary - Case=Loc|Gender=Fem|Number=Sing|Number[psor]=Plur|Person=3|Person[psor]=1 can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:54,229 - WARNING - machamp.data.machamp_vocabulary - Evident=Nfh|Mood=Ind|Number=Sing|Person=3|Polarity=Pos|Tense=Past|VerbForm=Fin|Voice=Pass can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:54,271 - WARNING - machamp.data.machamp_vocabulary - Case=Ins|Number=Sing|Person=3|PronType=Dem can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:54,280 - WARNING - machamp.data.machamp_vocabulary - Aspect=Hab|Mood=Pot|Number=Plur|Person=2|Polarity=Pos|Tense=Pres|VerbForm=Fin can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:54,287 - WARNING - machamp.data.machamp_vocabulary - Case=Gen|Number=Plur|Number[psor]=Plur|Person=1|Person[psor]=1 can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:54,320 - WARNING - machamp.data.machamp_vocabulary - Case=Ins|Number=Sing|Person=3|Polarity=Pos|VerbForm=Vnoun|Voice=Pass can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:54,333 - WARNING - machamp.data.machamp_vocabulary - Aspect=Hab|Mood=Ind|Polarity=Pos|Tense=Pres|VerbForm=Part can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:07:54,333 - INFO - machamp.readers.read_sequence - Stats UD-EWT (/content/dev.conllu):
2025-07-10 11:07:54,333 - INFO - machamp.readers.read_sequence - Lines:      291
2025-07-10 11:07:54,333 - INFO - machamp.readers.read_sequence - Words:      2,981
2025-07-10 11:07:54,333 - INFO - machamp.readers.read_sequence - Subwords:   9,043
2025-07-10 11:07:54,333 - INFO - machamp.readers.read_sequence - Unks:       130
2025-07-10 11:07:54,334 - INFO - machamp.readers.read_sequence - Pre-splits: 0
2025-07-10 11:07:54,334 - INFO - machamp.data.machamp_dataset - Done reading /content/dev.conllu (1.0s)

2025-07-10 11:07:54,549 - ERROR - STDERR - 
2025-07-10 11:07:54,549 - ERROR - STDERR - config.json:   0%|          | 0.00/668 [00:00<?, ?B/s]
2025-07-10 11:07:54,550 - ERROR - STDERR - 
2025-07-10 11:07:54,550 - ERROR - STDERR - config.json: 100%|##########| 668/668 [00:00<00:00, 5.38MB/s]
2025-07-10 11:07:57,160 - ERROR - STDERR - 
2025-07-10 11:07:57,160 - ERROR - STDERR - model.safetensors:   0%|          | 0.00/1.11G [00:00<?, ?B/s]
2025-07-10 11:07:57,357 - ERROR - STDERR - 
2025-07-10 11:07:57,358 - ERROR - STDERR - model.safetensors:   2%|1         | 21.0M/1.11G [00:00<00:10, 107MB/s]
2025-07-10 11:07:57,586 - ERROR - STDERR - 
2025-07-10 11:07:57,586 - ERROR - STDERR - model.safetensors:   4%|3         | 41.9M/1.11G [00:00<00:11, 97.3MB/s]
2025-07-10 11:07:57,722 - ERROR - STDERR - 
2025-07-10 11:07:57,722 - ERROR - STDERR - model.safetensors:   5%|4         | 52.4M/1.11G [00:00<00:11, 90.3MB/s]
2025-07-10 11:07:57,906 - ERROR - STDERR - 
2025-07-10 11:07:57,906 - ERROR - STDERR - model.safetensors:   7%|6         | 73.4M/1.11G [00:00<00:10, 99.8MB/s]
2025-07-10 11:07:58,059 - ERROR - STDERR - 
2025-07-10 11:07:58,059 - ERROR - STDERR - model.safetensors:   8%|8         | 94.4M/1.11G [00:00<00:09, 112MB/s]
2025-07-10 11:07:58,299 - ERROR - STDERR - 
2025-07-10 11:07:58,299 - ERROR - STDERR - model.safetensors:  10%|#         | 115M/1.11G [00:01<00:09, 101MB/s]
2025-07-10 11:07:58,505 - ERROR - STDERR - 
2025-07-10 11:07:58,506 - ERROR - STDERR - model.safetensors:  12%|#2        | 136M/1.11G [00:01<00:09, 102MB/s]
2025-07-10 11:07:58,613 - ERROR - STDERR - 
2025-07-10 11:07:58,613 - ERROR - STDERR - model.safetensors:  13%|#3        | 147M/1.11G [00:01<00:09, 101MB/s]
2025-07-10 11:07:58,809 - ERROR - STDERR - 
2025-07-10 11:07:58,809 - ERROR - STDERR - model.safetensors:  15%|#5        | 168M/1.11G [00:01<00:09, 103MB/s]
2025-07-10 11:07:58,936 - ERROR - STDERR - 
2025-07-10 11:07:58,936 - ERROR - STDERR - model.safetensors:  16%|#6        | 178M/1.11G [00:01<00:09, 97.9MB/s]
2025-07-10 11:07:59,159 - ERROR - STDERR - 
2025-07-10 11:07:59,159 - ERROR - STDERR - model.safetensors:  18%|#7        | 199M/1.11G [00:01<00:09, 96.5MB/s]
2025-07-10 11:07:59,262 - ERROR - STDERR - 
2025-07-10 11:07:59,262 - ERROR - STDERR - model.safetensors:  19%|#8        | 210M/1.11G [00:02<00:09, 97.6MB/s]
2025-07-10 11:07:59,439 - ERROR - STDERR - 
2025-07-10 11:07:59,439 - ERROR - STDERR - model.safetensors:  21%|##        | 231M/1.11G [00:02<00:08, 104MB/s]
2025-07-10 11:07:59,585 - ERROR - STDERR - 
2025-07-10 11:07:59,585 - ERROR - STDERR - model.safetensors:  23%|##2       | 252M/1.11G [00:02<00:07, 115MB/s]
2025-07-10 11:07:59,917 - ERROR - STDERR - 
2025-07-10 11:07:59,917 - ERROR - STDERR - model.safetensors:  24%|##4       | 273M/1.11G [00:02<00:09, 90.6MB/s]
2025-07-10 11:08:00,081 - ERROR - STDERR - 
2025-07-10 11:08:00,081 - ERROR - STDERR - model.safetensors:  26%|##6       | 294M/1.11G [00:02<00:08, 100MB/s]
2025-07-10 11:08:00,236 - ERROR - STDERR - 
2025-07-10 11:08:00,236 - ERROR - STDERR - model.safetensors:  28%|##8       | 315M/1.11G [00:03<00:07, 109MB/s]
2025-07-10 11:08:00,413 - ERROR - STDERR - 
2025-07-10 11:08:00,413 - ERROR - STDERR - model.safetensors:  30%|###       | 336M/1.11G [00:03<00:06, 112MB/s]
2025-07-10 11:08:00,700 - ERROR - STDERR - 
2025-07-10 11:08:00,701 - ERROR - STDERR - model.safetensors:  32%|###2      | 357M/1.11G [00:03<00:07, 96.0MB/s]
2025-07-10 11:08:00,933 - ERROR - STDERR - 
2025-07-10 11:08:00,933 - ERROR - STDERR - model.safetensors:  34%|###3      | 377M/1.11G [00:03<00:07, 94.2MB/s]
2025-07-10 11:08:01,057 - ERROR - STDERR - 
2025-07-10 11:08:01,057 - ERROR - STDERR - model.safetensors:  35%|###4      | 388M/1.11G [00:03<00:07, 92.3MB/s]
2025-07-10 11:08:01,243 - ERROR - STDERR - 
2025-07-10 11:08:01,243 - ERROR - STDERR - model.safetensors:  37%|###6      | 409M/1.11G [00:04<00:07, 98.3MB/s]
2025-07-10 11:08:01,389 - ERROR - STDERR - 
2025-07-10 11:08:01,389 - ERROR - STDERR - model.safetensors:  38%|###7      | 419M/1.11G [00:04<00:07, 91.7MB/s]
2025-07-10 11:08:01,579 - ERROR - STDERR - 
2025-07-10 11:08:01,579 - ERROR - STDERR - model.safetensors:  40%|###9      | 440M/1.11G [00:04<00:06, 97.7MB/s]
2025-07-10 11:08:01,716 - ERROR - STDERR - 
2025-07-10 11:08:01,716 - ERROR - STDERR - model.safetensors:  41%|####1     | 461M/1.11G [00:04<00:05, 111MB/s]
2025-07-10 11:08:01,979 - ERROR - STDERR - 
2025-07-10 11:08:01,980 - ERROR - STDERR - model.safetensors:  43%|####3     | 482M/1.11G [00:04<00:06, 98.5MB/s]
2025-07-10 11:08:02,134 - ERROR - STDERR - 
2025-07-10 11:08:02,134 - ERROR - STDERR - model.safetensors:  45%|####5     | 503M/1.11G [00:04<00:05, 108MB/s]
2025-07-10 11:08:02,345 - ERROR - STDERR - 
2025-07-10 11:08:02,345 - ERROR - STDERR - model.safetensors:  47%|####7     | 524M/1.11G [00:05<00:05, 105MB/s]
2025-07-10 11:08:02,577 - ERROR - STDERR - 
2025-07-10 11:08:02,578 - ERROR - STDERR - model.safetensors:  49%|####8     | 545M/1.11G [00:05<00:05, 100MB/s]
2025-07-10 11:08:02,774 - ERROR - STDERR - 
2025-07-10 11:08:02,774 - ERROR - STDERR - model.safetensors:  51%|#####     | 566M/1.11G [00:05<00:05, 102MB/s]
2025-07-10 11:08:02,904 - ERROR - STDERR - 
2025-07-10 11:08:02,904 - ERROR - STDERR - model.safetensors:  52%|#####1    | 577M/1.11G [00:05<00:05, 97.3MB/s]
2025-07-10 11:08:03,095 - ERROR - STDERR - 
2025-07-10 11:08:03,095 - ERROR - STDERR - model.safetensors:  54%|#####3    | 598M/1.11G [00:05<00:05, 101MB/s]
2025-07-10 11:08:03,257 - ERROR - STDERR - 
2025-07-10 11:08:03,257 - ERROR - STDERR - model.safetensors:  55%|#####4    | 608M/1.11G [00:06<00:05, 91.3MB/s]
2025-07-10 11:08:03,430 - ERROR - STDERR - 
2025-07-10 11:08:03,430 - ERROR - STDERR - model.safetensors:  57%|#####6    | 629M/1.11G [00:06<00:04, 100MB/s]
2025-07-10 11:08:03,641 - ERROR - STDERR - 
2025-07-10 11:08:03,641 - ERROR - STDERR - model.safetensors:  58%|#####8    | 650M/1.11G [00:06<00:04, 99.8MB/s]
2025-07-10 11:08:03,782 - ERROR - STDERR - 
2025-07-10 11:08:03,783 - ERROR - STDERR - model.safetensors:  59%|#####9    | 661M/1.11G [00:06<00:04, 93.5MB/s]
2025-07-10 11:08:03,951 - ERROR - STDERR - 
2025-07-10 11:08:03,951 - ERROR - STDERR - model.safetensors:  61%|######1   | 682M/1.11G [00:06<00:04, 103MB/s]
2025-07-10 11:08:04,155 - ERROR - STDERR - 
2025-07-10 11:08:04,155 - ERROR - STDERR - model.safetensors:  62%|######2   | 692M/1.11G [00:06<00:04, 85.3MB/s]
2025-07-10 11:08:04,289 - ERROR - STDERR - 
2025-07-10 11:08:04,289 - ERROR - STDERR - model.safetensors:  64%|######4   | 713M/1.11G [00:07<00:03, 102MB/s]
2025-07-10 11:08:04,445 - ERROR - STDERR - 
2025-07-10 11:08:04,445 - ERROR - STDERR - model.safetensors:  66%|######5   | 734M/1.11G [00:07<00:03, 112MB/s]
2025-07-10 11:08:04,814 - ERROR - STDERR - 
2025-07-10 11:08:04,814 - ERROR - STDERR - model.safetensors:  68%|######7   | 755M/1.11G [00:07<00:04, 84.7MB/s]
2025-07-10 11:08:04,998 - ERROR - STDERR - 
2025-07-10 11:08:04,999 - ERROR - STDERR - model.safetensors:  70%|######9   | 776M/1.11G [00:07<00:03, 92.2MB/s]
2025-07-10 11:08:05,113 - ERROR - STDERR - 
2025-07-10 11:08:05,113 - ERROR - STDERR - model.safetensors:  71%|#######   | 786M/1.11G [00:07<00:03, 92.1MB/s]
2025-07-10 11:08:05,301 - ERROR - STDERR - 
2025-07-10 11:08:05,301 - ERROR - STDERR - model.safetensors:  73%|#######2  | 807M/1.11G [00:08<00:03, 98.1MB/s]
2025-07-10 11:08:05,417 - ERROR - STDERR - 
2025-07-10 11:08:05,417 - ERROR - STDERR - model.safetensors:  73%|#######3  | 818M/1.11G [00:08<00:03, 96.3MB/s]
2025-07-10 11:08:05,651 - ERROR - STDERR - 
2025-07-10 11:08:05,651 - ERROR - STDERR - model.safetensors:  75%|#######5  | 839M/1.11G [00:08<00:02, 93.9MB/s]
2025-07-10 11:08:05,952 - ERROR - STDERR - 
2025-07-10 11:08:05,952 - ERROR - STDERR - model.safetensors:  76%|#######6  | 849M/1.11G [00:08<00:03, 69.6MB/s]
2025-07-10 11:08:06,137 - ERROR - STDERR - 
2025-07-10 11:08:06,137 - ERROR - STDERR - model.safetensors:  78%|#######8  | 870M/1.11G [00:08<00:02, 81.2MB/s]
2025-07-10 11:08:06,293 - ERROR - STDERR - 
2025-07-10 11:08:06,293 - ERROR - STDERR - model.safetensors:  79%|#######9  | 881M/1.11G [00:09<00:02, 77.9MB/s]
2025-07-10 11:08:06,448 - ERROR - STDERR - 
2025-07-10 11:08:06,449 - ERROR - STDERR - model.safetensors:  81%|########1 | 902M/1.11G [00:09<00:02, 92.4MB/s]
2025-07-10 11:08:06,599 - ERROR - STDERR - 
2025-07-10 11:08:06,599 - ERROR - STDERR - model.safetensors:  82%|########1 | 912M/1.11G [00:09<00:02, 86.5MB/s]
2025-07-10 11:08:06,896 - ERROR - STDERR - 
2025-07-10 11:08:06,896 - ERROR - STDERR - model.safetensors:  84%|########3 | 933M/1.11G [00:09<00:02, 79.7MB/s]
2025-07-10 11:08:07,051 - ERROR - STDERR - 
2025-07-10 11:08:07,051 - ERROR - STDERR - model.safetensors:  86%|########5 | 954M/1.11G [00:09<00:01, 93.1MB/s]
2025-07-10 11:08:07,210 - ERROR - STDERR - 
2025-07-10 11:08:07,210 - ERROR - STDERR - model.safetensors:  87%|########6 | 965M/1.11G [00:10<00:01, 86.1MB/s]
2025-07-10 11:08:07,393 - ERROR - STDERR - 
2025-07-10 11:08:07,393 - ERROR - STDERR - model.safetensors:  89%|########8 | 986M/1.11G [00:10<00:01, 94.7MB/s]
2025-07-10 11:08:07,513 - ERROR - STDERR - 
2025-07-10 11:08:07,513 - ERROR - STDERR - model.safetensors:  89%|########9 | 996M/1.11G [00:10<00:01, 93.0MB/s]
2025-07-10 11:08:07,766 - ERROR - STDERR - 
2025-07-10 11:08:07,766 - ERROR - STDERR - model.safetensors:  91%|#########1| 1.02G/1.11G [00:10<00:01, 89.0MB/s]
2025-07-10 11:08:07,898 - ERROR - STDERR - 
2025-07-10 11:08:07,899 - ERROR - STDERR - model.safetensors:  92%|#########2| 1.03G/1.11G [00:10<00:00, 86.7MB/s]
2025-07-10 11:08:08,105 - ERROR - STDERR - 
2025-07-10 11:08:08,105 - ERROR - STDERR - model.safetensors:  94%|#########4| 1.05G/1.11G [00:10<00:00, 91.8MB/s]
2025-07-10 11:08:08,208 - ERROR - STDERR - 
2025-07-10 11:08:08,208 - ERROR - STDERR - model.safetensors:  95%|#########5| 1.06G/1.11G [00:11<00:00, 93.7MB/s]
2025-07-10 11:08:08,372 - ERROR - STDERR - 
2025-07-10 11:08:08,372 - ERROR - STDERR - model.safetensors:  97%|#########7| 1.08G/1.11G [00:11<00:00, 104MB/s]
2025-07-10 11:08:08,508 - ERROR - STDERR - 
2025-07-10 11:08:08,508 - ERROR - STDERR - model.safetensors:  98%|#########7| 1.09G/1.11G [00:11<00:00, 97.0MB/s]
2025-07-10 11:08:08,668 - ERROR - STDERR - 
2025-07-10 11:08:08,668 - ERROR - STDERR - model.safetensors: 100%|#########9| 1.11G/1.11G [00:11<00:00, 107MB/s]
2025-07-10 11:08:08,681 - ERROR - STDERR - 
2025-07-10 11:08:08,681 - ERROR - STDERR - model.safetensors: 100%|##########| 1.11G/1.11G [00:11<00:00, 96.6MB/s]
2025-07-10 11:08:09,311 - INFO - machamp.model.machamp - Overview of the torch model: 
2025-07-10 11:08:09,311 - INFO - machamp.model.machamp - MachampModel(
  (mlm): XLMRobertaModel(
    (embeddings): XLMRobertaEmbeddings(
      (word_embeddings): Embedding(250002, 768, padding_idx=1)
      (position_embeddings): Embedding(514, 768, padding_idx=1)
      (token_type_embeddings): Embedding(1, 768)
      (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
      (dropout): Dropout(p=0.1, inplace=False)
    )
    (encoder): XLMRobertaEncoder(
      (layer): ModuleList(
        (0-11): 12 x XLMRobertaLayer(
          (attention): XLMRobertaAttention(
            (self): XLMRobertaSdpaSelfAttention(
              (query): Linear(in_features=768, out_features=768, bias=True)
              (key): Linear(in_features=768, out_features=768, bias=True)
              (value): Linear(in_features=768, out_features=768, bias=True)
              (dropout): Dropout(p=0.1, inplace=False)
            )
            (output): XLMRobertaSelfOutput(
              (dense): Linear(in_features=768, out_features=768, bias=True)
              (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
              (dropout): Dropout(p=0.1, inplace=False)
            )
          )
          (intermediate): XLMRobertaIntermediate(
            (dense): Linear(in_features=768, out_features=3072, bias=True)
            (intermediate_act_fn): GELUActivation()
          )
          (output): XLMRobertaOutput(
            (dense): Linear(in_features=3072, out_features=768, bias=True)
            (LayerNorm): LayerNorm((768,), eps=1e-05, elementwise_affine=True)
            (dropout): Dropout(p=0.1, inplace=False)
          )
        )
      )
    )
    (pooler): XLMRobertaPooler(
      (dense): Linear(in_features=768, out_features=768, bias=True)
      (activation): Tanh()
    )
  )
  (dropout): Dropout(p=0.2, inplace=False)
  (decoders): ModuleDict(
    (dependency): MachampDepDecoder(
      (head_arc_feedforward): Linear(in_features=768, out_features=768, bias=True)
      (child_arc_feedforward): Linear(in_features=768, out_features=768, bias=True)
      (arc_attention): BilinearMatrixAttention()
      (head_tag_feedforward): Linear(in_features=768, out_features=256, bias=True)
      (child_tag_feedforward): Linear(in_features=768, out_features=256, bias=True)
      (tag_bilinear): Bilinear(in1_features=256, in2_features=256, out_features=38, bias=True)
      (decoder_dropout): Dropout(p=0.0, inplace=False)
    )
    (lemma): MachampSeqDecoder(
      (hidden_to_label): Linear(in_features=768, out_features=252, bias=True)
      (loss_function): CrossEntropyLoss()
      (decoder_dropout): Dropout(p=0.0, inplace=False)
    )
    (morph): MachampSeqDecoder(
      (hidden_to_label): Linear(in_features=768, out_features=480, bias=True)
      (loss_function): CrossEntropyLoss()
      (decoder_dropout): Dropout(p=0.0, inplace=False)
    )
    (upos): MachampSeqDecoder(
      (hidden_to_label): Linear(in_features=768, out_features=17, bias=True)
      (loss_function): CrossEntropyLoss()
      (decoder_dropout): Dropout(p=0.0, inplace=False)
    )
    (xpos): MachampSeqDecoder(
      (hidden_to_label): Linear(in_features=768, out_features=35, bias=True)
      (loss_function): CrossEntropyLoss()
      (decoder_dropout): Dropout(p=0.0, inplace=False)
    )
  )
  (scalars): ModuleDict(
    (dependency): ScalarMix()
    (lemma): ScalarMix()
    (morph): ScalarMix()
    (upos): ScalarMix()
    (xpos): ScalarMix()
  )
)
2025-07-10 11:08:09,348 - INFO - machamp.utils.myutils - Done constructing parameter groups.
2025-07-10 11:08:09,348 - INFO - machamp.utils.myutils - Group 0: ['mlm.encoder.layer.4.attention.self.query.bias', 'mlm.encoder.layer.0.attention.self.key.bias', 'mlm.encoder.layer.9.intermediate.dense.weight', 'mlm.encoder.layer.6.attention.self.value.weight', 'mlm.encoder.layer.3.intermediate.dense.bias', 'mlm.encoder.layer.3.attention.output.dense.bias', 'mlm.encoder.layer.3.attention.self.query.bias', 'mlm.encoder.layer.11.attention.self.query.weight', 'mlm.encoder.layer.7.attention.output.dense.weight', 'mlm.encoder.layer.7.output.LayerNorm.bias', 'mlm.encoder.layer.9.attention.output.LayerNorm.weight', 'mlm.encoder.layer.11.attention.output.dense.bias', 'mlm.encoder.layer.9.attention.output.dense.weight', 'mlm.encoder.layer.1.attention.output.LayerNorm.bias', 'mlm.encoder.layer.4.attention.output.dense.bias', 'mlm.encoder.layer.7.attention.self.value.weight', 'mlm.encoder.layer.9.attention.output.dense.bias', 'mlm.encoder.layer.6.output.dense.weight', 'mlm.encoder.layer.4.intermediate.dense.weight', 'mlm.encoder.layer.2.intermediate.dense.bias', 'mlm.encoder.layer.2.attention.self.query.weight', 'mlm.encoder.layer.3.attention.self.query.weight', 'mlm.encoder.layer.0.intermediate.dense.bias', 'mlm.encoder.layer.9.intermediate.dense.bias', 'mlm.encoder.layer.0.output.dense.weight', 'mlm.encoder.layer.10.attention.self.key.bias', 'mlm.encoder.layer.3.attention.output.LayerNorm.bias', 'mlm.encoder.layer.5.attention.self.value.weight', 'mlm.encoder.layer.11.attention.self.key.bias', 'mlm.encoder.layer.4.attention.output.dense.weight', 'mlm.encoder.layer.0.attention.output.dense.bias', 'mlm.encoder.layer.1.output.dense.weight', 'mlm.encoder.layer.7.attention.output.dense.bias', 'mlm.encoder.layer.0.attention.output.dense.weight', 'mlm.encoder.layer.8.attention.self.query.weight', 'mlm.encoder.layer.5.attention.output.dense.weight', 'mlm.encoder.layer.0.attention.self.query.weight', 'mlm.encoder.layer.10.attention.self.key.weight', 'mlm.encoder.layer.11.output.LayerNorm.weight', 'mlm.encoder.layer.2.attention.self.value.weight', 'mlm.encoder.layer.0.attention.output.LayerNorm.bias', 'mlm.encoder.layer.2.attention.self.query.bias', 'mlm.encoder.layer.7.attention.output.LayerNorm.weight', 'mlm.encoder.layer.1.attention.self.query.bias', 'mlm.encoder.layer.5.attention.output.LayerNorm.bias', 'mlm.encoder.layer.8.attention.output.LayerNorm.bias', 'mlm.encoder.layer.9.attention.self.query.weight', 'mlm.encoder.layer.11.attention.output.LayerNorm.bias', 'mlm.embeddings.LayerNorm.weight', 'mlm.encoder.layer.1.attention.self.value.weight', 'mlm.encoder.layer.3.attention.self.key.weight', 'mlm.encoder.layer.3.attention.output.LayerNorm.weight', 'mlm.encoder.layer.7.attention.self.key.bias', 'mlm.encoder.layer.4.attention.self.query.weight', 'mlm.encoder.layer.4.output.LayerNorm.bias', 'mlm.encoder.layer.3.intermediate.dense.weight', 'mlm.encoder.layer.5.attention.self.query.bias', 'mlm.embeddings.LayerNorm.bias', 'mlm.encoder.layer.3.output.dense.bias', 'mlm.encoder.layer.7.attention.self.query.bias', 'mlm.embeddings.word_embeddings.weight', 'mlm.encoder.layer.1.attention.self.value.bias', 'mlm.encoder.layer.5.attention.self.key.bias', 'mlm.encoder.layer.6.attention.output.LayerNorm.weight', 'mlm.encoder.layer.5.attention.self.query.weight', 'mlm.encoder.layer.6.attention.self.value.bias', 'mlm.encoder.layer.4.attention.self.key.weight', 'mlm.encoder.layer.10.attention.output.LayerNorm.weight', 'mlm.encoder.layer.10.output.dense.bias', 'mlm.encoder.layer.11.attention.output.dense.weight', 'mlm.encoder.layer.9.attention.self.key.weight', 'mlm.encoder.layer.10.attention.self.value.bias', 'mlm.encoder.layer.7.attention.output.LayerNorm.bias', 'mlm.encoder.layer.11.output.LayerNorm.bias', 'mlm.encoder.layer.1.intermediate.dense.weight', 'mlm.encoder.layer.2.attention.output.dense.bias', 'mlm.encoder.layer.4.attention.output.LayerNorm.weight', 'mlm.encoder.layer.5.intermediate.dense.bias', 'mlm.encoder.layer.6.output.dense.bias', 'mlm.encoder.layer.2.output.LayerNorm.weight', 'mlm.encoder.layer.1.attention.output.dense.bias', 'mlm.encoder.layer.2.attention.self.value.bias', 'mlm.encoder.layer.2.output.dense.weight', 'mlm.encoder.layer.11.attention.self.query.bias', 'mlm.encoder.layer.6.attention.output.dense.bias', 'mlm.encoder.layer.1.output.LayerNorm.weight', 'mlm.encoder.layer.10.attention.output.LayerNorm.bias', 'mlm.encoder.layer.11.attention.self.key.weight', 'mlm.encoder.layer.6.attention.self.key.bias', 'mlm.encoder.layer.4.output.LayerNorm.weight', 'mlm.encoder.layer.10.attention.self.value.weight', 'mlm.encoder.layer.0.output.LayerNorm.weight', 'mlm.encoder.layer.4.attention.self.key.bias', 'mlm.encoder.layer.0.intermediate.dense.weight', 'mlm.encoder.layer.1.output.dense.bias', 'mlm.encoder.layer.4.attention.self.value.bias', 'mlm.encoder.layer.5.output.LayerNorm.weight', 'mlm.encoder.layer.6.attention.self.query.bias', 'mlm.encoder.layer.2.output.LayerNorm.bias', 'mlm.encoder.layer.10.output.LayerNorm.weight', 'mlm.encoder.layer.6.attention.output.LayerNorm.bias', 'mlm.encoder.layer.4.output.dense.weight', 'mlm.encoder.layer.7.attention.self.value.bias', 'mlm.encoder.layer.10.intermediate.dense.bias', 'mlm.encoder.layer.8.output.LayerNorm.bias', 'mlm.encoder.layer.9.attention.self.key.bias', 'mlm.encoder.layer.9.output.dense.weight', 'mlm.encoder.layer.3.attention.output.dense.weight', 'mlm.encoder.layer.0.attention.output.LayerNorm.weight', 'mlm.encoder.layer.9.attention.output.LayerNorm.bias', 'mlm.encoder.layer.5.output.dense.weight', 'mlm.encoder.layer.8.attention.self.value.bias', 'mlm.encoder.layer.5.attention.self.value.bias', 'mlm.encoder.layer.0.attention.self.value.weight', 'mlm.encoder.layer.3.output.LayerNorm.bias', 'mlm.encoder.layer.5.attention.self.key.weight', 'mlm.encoder.layer.7.attention.self.query.weight', 'mlm.encoder.layer.10.intermediate.dense.weight', 'mlm.embeddings.position_embeddings.weight', 'mlm.encoder.layer.1.output.LayerNorm.bias', 'mlm.encoder.layer.2.attention.self.key.bias', 'mlm.encoder.layer.2.attention.output.LayerNorm.weight', 'mlm.encoder.layer.4.attention.output.LayerNorm.bias', 'mlm.encoder.layer.6.output.LayerNorm.weight', 'mlm.encoder.layer.3.attention.self.key.bias', 'mlm.encoder.layer.0.attention.self.query.bias', 'mlm.encoder.layer.4.intermediate.dense.bias', 'mlm.encoder.layer.8.attention.self.query.bias', 'mlm.encoder.layer.8.intermediate.dense.weight', 'mlm.encoder.layer.9.output.LayerNorm.bias', 'mlm.encoder.layer.1.attention.self.key.bias', 'mlm.encoder.layer.5.intermediate.dense.weight', 'mlm.encoder.layer.0.output.LayerNorm.bias', 'mlm.encoder.layer.8.output.dense.weight', 'mlm.encoder.layer.0.output.dense.bias', 'mlm.encoder.layer.6.output.LayerNorm.bias', 'mlm.encoder.layer.3.attention.self.value.weight', 'mlm.encoder.layer.5.output.LayerNorm.bias', 'mlm.encoder.layer.10.output.dense.weight', 'mlm.encoder.layer.8.attention.self.key.bias', 'mlm.encoder.layer.7.attention.self.key.weight', 'mlm.encoder.layer.2.intermediate.dense.weight', 'mlm.encoder.layer.9.output.LayerNorm.weight', 'mlm.encoder.layer.11.attention.self.value.bias', 'mlm.encoder.layer.8.output.dense.bias', 'mlm.encoder.layer.3.output.LayerNorm.weight', 'mlm.encoder.layer.6.attention.self.key.weight', 'mlm.encoder.layer.10.output.LayerNorm.bias', 'mlm.encoder.layer.9.attention.self.query.bias', 'mlm.encoder.layer.8.attention.output.dense.bias', 'mlm.encoder.layer.2.attention.output.LayerNorm.bias', 'mlm.encoder.layer.5.attention.output.dense.bias', 'mlm.encoder.layer.8.attention.output.LayerNorm.weight', 'mlm.encoder.layer.11.intermediate.dense.bias', 'mlm.encoder.layer.2.output.dense.bias', 'mlm.encoder.layer.7.output.LayerNorm.weight', 'mlm.encoder.layer.8.intermediate.dense.bias', 'mlm.encoder.layer.11.attention.self.value.weight', 'mlm.encoder.layer.9.attention.self.value.bias', 'mlm.encoder.layer.5.attention.output.LayerNorm.weight', 'mlm.encoder.layer.7.output.dense.weight', 'mlm.encoder.layer.8.output.LayerNorm.weight', 'mlm.encoder.layer.6.attention.output.dense.weight', 'mlm.encoder.layer.8.attention.self.key.weight', 'mlm.embeddings.token_type_embeddings.weight', 'mlm.encoder.layer.9.attention.self.value.weight', 'mlm.encoder.layer.11.output.dense.weight', 'mlm.encoder.layer.0.attention.self.key.weight', 'mlm.encoder.layer.1.attention.self.query.weight', 'mlm.encoder.layer.10.attention.output.dense.bias', 'mlm.encoder.layer.8.attention.self.value.weight', 'mlm.encoder.layer.10.attention.output.dense.weight', 'mlm.encoder.layer.11.attention.output.LayerNorm.weight', 'mlm.encoder.layer.1.attention.self.key.weight', 'mlm.encoder.layer.0.attention.self.value.bias', 'mlm.pooler.dense.bias', 'mlm.encoder.layer.1.intermediate.dense.bias', 'mlm.encoder.layer.11.output.dense.bias', 'mlm.encoder.layer.5.output.dense.bias', 'mlm.encoder.layer.8.attention.output.dense.weight', 'mlm.encoder.layer.3.attention.self.value.bias', 'mlm.encoder.layer.7.output.dense.bias', 'mlm.encoder.layer.9.output.dense.bias', 'mlm.encoder.layer.10.attention.self.query.bias', 'mlm.encoder.layer.6.attention.self.query.weight', 'mlm.encoder.layer.4.output.dense.bias', 'mlm.encoder.layer.1.attention.output.dense.weight', 'mlm.encoder.layer.1.attention.output.LayerNorm.weight', 'mlm.encoder.layer.6.intermediate.dense.bias', 'mlm.encoder.layer.7.intermediate.dense.weight', 'mlm.encoder.layer.7.intermediate.dense.bias', 'mlm.encoder.layer.11.intermediate.dense.weight', 'mlm.pooler.dense.weight', 'mlm.encoder.layer.2.attention.self.key.weight', 'mlm.encoder.layer.3.output.dense.weight', 'mlm.encoder.layer.4.attention.self.value.weight', 'mlm.encoder.layer.6.intermediate.dense.weight', 'mlm.encoder.layer.2.attention.output.dense.weight', 'mlm.encoder.layer.10.attention.self.query.weight'], {}
2025-07-10 11:08:09,348 - INFO - machamp.utils.myutils - Group 1: ['scalars.lemma.scalar_parameters', 'scalars.upos.gamma', 'scalars.xpos.gamma', 'scalars.upos.scalar_parameters', 'decoders.dependency.child_arc_feedforward.weight', 'decoders.xpos.hidden_to_label.weight', 'decoders.dependency.child_tag_feedforward.weight', 'decoders.upos.hidden_to_label.bias', 'decoders.dependency.tag_bilinear.weight', 'decoders.lemma.hidden_to_label.weight', 'decoders.lemma.hidden_to_label.bias', 'decoders.upos.hidden_to_label.weight', 'decoders.dependency.head_tag_feedforward.weight', 'decoders.morph.hidden_to_label.bias', 'scalars.dependency.gamma', 'scalars.xpos.scalar_parameters', 'decoders.morph.hidden_to_label.weight', 'decoders.dependency.head_arc_feedforward.weight', 'scalars.morph.gamma', 'decoders.dependency.arc_attention._bias', 'decoders.dependency._head_sentinel', 'decoders.dependency.head_arc_feedforward.bias', 'decoders.dependency.child_arc_feedforward.bias', 'decoders.dependency.arc_attention._weight_matrix', 'decoders.dependency.tag_bilinear.bias', 'scalars.lemma.gamma', 'decoders.dependency.child_tag_feedforward.bias', 'scalars.dependency.scalar_parameters', 'scalars.morph.scalar_parameters', 'decoders.xpos.hidden_to_label.bias', 'decoders.dependency.head_tag_feedforward.bias'], {}
2025-07-10 11:08:09,348 - INFO - machamp.utils.myutils - Group 2: [], {}
2025-07-10 11:08:09,348 - INFO - machamp.utils.myutils - Number of trainable parameters: 283304012
2025-07-10 11:08:09,349 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing. Training only the decoder heads, not the MLM.
2025-07-10 11:08:09,351 - INFO - machamp.model.trainer - MaChAmp succesfully initialized in 24.0s
2025-07-10 11:08:09,352 - INFO - machamp.model.trainer - 

          CHAMP!                   %%#/   
 ###/%(        CHAMP!            ,#//(%%#( 
#%%/((/(            . , ,       .%%#####%(
#(/(((###(        ////////%     *//((((((/
  ,((((//        #%(%#/#%%/          &####
     ###### *#%%@&((//&@#%&(%%#%%%%#%%####
     ,####(%%###%#&&&#((&&(((%%%%########(
       ///((%%%##/%%######((%%%%%%%%((((//
          %%%%%##((/%(///%%##%%%%%%%##    
        /%%####%%%%%%#%%#/(%#(((%%%%%&.   
       &#%/#%%###%%%##%%%((###(//(####(   
     @&%%#////(########%######((/(////(   
       #////     ((((##(((((////,         
                   */(//(((/**,.          
                  #,**%#%,*,,,*(          
               /%%%%%/,,,,,(%%%%((        
              &%%%#(((/*  /(#%%%#(/       
              %%#(((//     ,((###(/       
             .%(/(/,         /(###/(      
              %(((//          (/((///     
               ((/,             /(//      
             %%%#((,            &%%%#     
         /#(#%(/,             /###*##*    

2025-07-10 11:08:09,352 - INFO - machamp.model.trainer - starting training...
2025-07-10 11:08:09,352 - INFO - machamp.model.trainer - Epoch 1/60: training
2025-07-10 11:08:18,006 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:08:18,006 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 6.25e-07
2025-07-10 11:08:18,006 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 6.25e-07
2025-07-10 11:08:18,006 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 2.3750000000000003e-07
2025-07-10 11:08:18,006 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:08:18,006 - INFO - machamp.model.trainer - Epoch 1: evaluating on dev
2025-07-10 11:08:19,518 - INFO - machamp.model.callback - epoch       : 1/60
2025-07-10 11:08:19,518 - INFO - machamp.model.callback - best_epoch  : 1
2025-07-10 11:08:19,518 - INFO - machamp.model.callback - max_gpu_mem : 1.8494
2025-07-10 11:08:19,518 - INFO - machamp.model.callback - cur_ram     : 2.1156
2025-07-10 11:08:19,518 - INFO - machamp.model.callback - time_epoch  : 0:00:10
2025-07-10 11:08:19,518 - INFO - machamp.model.callback - time_total  : 0:00:10
2025-07-10 11:08:19,518 - INFO - machamp.model.callback - 
2025-07-10 11:08:19,518 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:08:19,518 - INFO - machamp.model.callback - Best (1)                                                   
2025-07-10 11:08:19,518 - INFO - machamp.model.callback - dependency_las     6.5414   4.4852       0.0620     0.1234 
2025-07-10 11:08:19,518 - INFO - machamp.model.callback - lemma_accuracy     4.5302   4.4153       0.0032     0.0054 
2025-07-10 11:08:19,518 - INFO - machamp.model.callback - morph_accuracy     6.0533   5.9711       0.0143     0.0265 
2025-07-10 11:08:19,518 - INFO - machamp.model.callback - upos_accuracy      1.3574   1.3181       0.1484     0.2794 
2025-07-10 11:08:19,519 - INFO - machamp.model.callback - xpos_accuracy      1.7460   1.6948       0.0230     0.0282 
2025-07-10 11:08:19,519 - INFO - machamp.model.callback - sum               20.2282  17.8845       0.2509     0.4629 
2025-07-10 11:08:19,519 - INFO - machamp.model.callback - Epoch 1                                                    
2025-07-10 11:08:19,519 - INFO - machamp.model.callback - dependency_las     6.5414   4.4852       0.0620     0.1234 
2025-07-10 11:08:19,519 - INFO - machamp.model.callback - lemma_accuracy     4.5302   4.4153       0.0032     0.0054 
2025-07-10 11:08:19,519 - INFO - machamp.model.callback - morph_accuracy     6.0533   5.9711       0.0143     0.0265 
2025-07-10 11:08:19,519 - INFO - machamp.model.callback - upos_accuracy      1.3574   1.3181       0.1484     0.2794 
2025-07-10 11:08:19,519 - INFO - machamp.model.callback - xpos_accuracy      1.7460   1.6948       0.0230     0.0282 
2025-07-10 11:08:19,519 - INFO - machamp.model.callback - sum               20.2282  17.8845       0.2509     0.4629 
2025-07-10 11:08:19,534 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
││                             ▘                             │
││                                                           │
││                                                           │
││                                                           │
││                                                           │ 0.2
││                                                           │
││                                                           │
││                                                           │
││                                                           │
││                             ▖                             │
││                                                           │
││                                                           │ 0.1
││                                                           │
││                                                           │
││                                                           │
││                             ▘                             │
││                             ▖                             │
└────────────────────────────────────────────────────────────┘
0.0            0.5            1.0            1.5
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:08:19,534 - INFO - machamp.model.callback - Performance of 0.4629 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_1.pt
2025-07-10 11:08:21,280 - INFO - machamp.model.trainer - 
2025-07-10 11:08:21,280 - INFO - machamp.model.trainer - Epoch 2/60: training
2025-07-10 11:08:47,922 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:08:47,922 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.7197174922600621e-06
2025-07-10 11:08:47,922 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.7197174922600621e-06
2025-07-10 11:08:47,922 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 6.534926470588237e-07
2025-07-10 11:08:47,922 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:08:47,922 - INFO - machamp.model.trainer - Epoch 2: evaluating on dev
2025-07-10 11:08:49,557 - INFO - machamp.model.callback - epoch       : 2/60
2025-07-10 11:08:49,557 - INFO - machamp.model.callback - best_epoch  : 2
2025-07-10 11:08:49,557 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:08:49,557 - INFO - machamp.model.callback - cur_ram     : 2.1356
2025-07-10 11:08:49,558 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:08:49,558 - INFO - machamp.model.callback - time_total  : 0:00:40
2025-07-10 11:08:49,558 - INFO - machamp.model.callback - 
2025-07-10 11:08:49,558 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:08:49,558 - INFO - machamp.model.callback - Best (2)                                                   
2025-07-10 11:08:49,558 - INFO - machamp.model.callback - dependency_las     4.9742   4.2129       0.1078     0.1422 
2025-07-10 11:08:49,558 - INFO - machamp.model.callback - lemma_accuracy     4.4095   4.3436       0.0060     0.0074 
2025-07-10 11:08:49,558 - INFO - machamp.model.callback - morph_accuracy     5.9460   5.9191       0.0275     0.0312 
2025-07-10 11:08:49,558 - INFO - machamp.model.callback - upos_accuracy      1.3175   1.2996       0.2425     0.3241 
2025-07-10 11:08:49,558 - INFO - machamp.model.callback - xpos_accuracy      1.6832   1.6569       0.0546     0.0500 
2025-07-10 11:08:49,558 - INFO - machamp.model.callback - sum               18.3304  17.4320       0.4385     0.5548 
2025-07-10 11:08:49,558 - INFO - machamp.model.callback - Epoch 2                                                    
2025-07-10 11:08:49,559 - INFO - machamp.model.callback - dependency_las     4.9742   4.2129       0.1078     0.1422 
2025-07-10 11:08:49,559 - INFO - machamp.model.callback - lemma_accuracy     4.4095   4.3436       0.0060     0.0074 
2025-07-10 11:08:49,559 - INFO - machamp.model.callback - morph_accuracy     5.9460   5.9191       0.0275     0.0312 
2025-07-10 11:08:49,559 - INFO - machamp.model.callback - upos_accuracy      1.3175   1.2996       0.2425     0.3241 
2025-07-10 11:08:49,559 - INFO - machamp.model.callback - xpos_accuracy      1.6832   1.6569       0.0546     0.0500 
2025-07-10 11:08:49,559 - INFO - machamp.model.callback - sum               18.3304  17.4320       0.4385     0.5548 
2025-07-10 11:08:49,566 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                                         ▄▄▄▄▄▄▄▄▄▄▄▄▞▀▀▀▀▀▀│
│               ▄▄▄▄▄▄▄▄▄▄▄▄▄▀▀▀▀▀▀▀▀▀▀▀▀▀                   │ 0.3
│▄▄▀▀▀▀▀▀▀▀▀▀▀▀▀                                             │
│                                                            │
│                                                            │
│                                                            │
│                                                            │ 0.2
│                                                            │
│                                                            │
│                                       ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄│
│▄▄▄▄▄▄▄▄▞▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀                     │
│                                                            │ 0.1
│                                                            │
│                                                            │
│                             ▗▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▀▀▀▀│
│▄▄▄▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▘▄▄▄▄▄▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀│
│▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄│
└────────────────────────────────────────────────────────────┘
1.0                           1.5                          2.0
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:08:49,566 - INFO - machamp.model.callback - Performance of 0.5548 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_2.pt
2025-07-10 11:08:51,372 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_1.pt
2025-07-10 11:08:51,514 - INFO - machamp.model.trainer - 
2025-07-10 11:08:51,514 - INFO - machamp.model.trainer - Epoch 3/60: training
2025-07-10 11:09:18,290 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:09:18,290 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 2.814434984520124e-06
2025-07-10 11:09:18,291 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 2.814434984520124e-06
2025-07-10 11:09:18,291 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 1.0694852941176473e-06
2025-07-10 11:09:18,291 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:09:18,291 - INFO - machamp.model.trainer - Epoch 3: evaluating on dev
2025-07-10 11:09:19,966 - INFO - machamp.model.callback - epoch       : 3/60
2025-07-10 11:09:19,966 - INFO - machamp.model.callback - best_epoch  : 3
2025-07-10 11:09:19,966 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:09:19,966 - INFO - machamp.model.callback - cur_ram     : 2.1236
2025-07-10 11:09:19,966 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:09:19,966 - INFO - machamp.model.callback - time_total  : 0:01:10
2025-07-10 11:09:19,966 - INFO - machamp.model.callback - 
2025-07-10 11:09:19,966 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:09:19,966 - INFO - machamp.model.callback - Best (3)                                                   
2025-07-10 11:09:19,966 - INFO - machamp.model.callback - dependency_las     4.6147   3.8996       0.1234     0.1748 
2025-07-10 11:09:19,966 - INFO - machamp.model.callback - lemma_accuracy     4.2996   4.1962       0.0114     0.0148 
2025-07-10 11:09:19,966 - INFO - machamp.model.callback - morph_accuracy     5.8731   5.8118       0.0372     0.0433 
2025-07-10 11:09:19,966 - INFO - machamp.model.callback - upos_accuracy      1.2899   1.2633       0.3057     0.3677 
2025-07-10 11:09:19,966 - INFO - machamp.model.callback - xpos_accuracy      1.6260   1.5786       0.1216     0.1875 
2025-07-10 11:09:19,966 - INFO - machamp.model.callback - sum               17.7033  16.7495       0.5992     0.7880 
2025-07-10 11:09:19,966 - INFO - machamp.model.callback - Epoch 3                                                    
2025-07-10 11:09:19,966 - INFO - machamp.model.callback - dependency_las     4.6147   3.8996       0.1234     0.1748 
2025-07-10 11:09:19,966 - INFO - machamp.model.callback - lemma_accuracy     4.2996   4.1962       0.0114     0.0148 
2025-07-10 11:09:19,967 - INFO - machamp.model.callback - morph_accuracy     5.8731   5.8118       0.0372     0.0433 
2025-07-10 11:09:19,967 - INFO - machamp.model.callback - upos_accuracy      1.2899   1.2633       0.3057     0.3677 
2025-07-10 11:09:19,967 - INFO - machamp.model.callback - xpos_accuracy      1.6260   1.5786       0.1216     0.1875 
2025-07-10 11:09:19,967 - INFO - machamp.model.callback - sum               17.7033  16.7495       0.5992     0.7880 
2025-07-10 11:09:19,974 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                                                ▗▄▄▄▄▄▄▄▀▀▀▀│
│                                 ▗▄▄▄▄▄▄▄▀▀▀▀▀▀▀▘           │
│                   ▄▄▄▄▄▄▄▞▀▀▀▀▀▀▘                          │
│    ▗▄▄▄▄▄▄▞▀▀▀▀▀▀▀                                         │ 0.3
│▀▀▀▀▘                                                       │
│                                                            │
│                                                            │
│                                                            │ 0.2
│                                                         ▄▄▞│
│                                        ▗▄▄▄▄▄▄▄▄▄▞▀▗▄▞▀▀▀▀▀│
│             ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▞▀▀▀▀▀▀▀▀▀▘      ▗▄▄▀▀▘       │
│▀▀▀▀▀▀▀▀▀▀▀▀▀                              ▄▄▀▀▘            │
│                                      ▄▄▞▀▀                 │ 0.1
│                                 ▗▄▞▀▀                      │
│                     ▄▄▄▄▄▄▄▄▄▄▀▀▘                          │
│▄▄▄▄▄▄▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▄▄▄▄▄▄▄▄▄▄▄▄▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀│
│▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▞▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀│
└────────────────────────────────────────────────────────────┘
 1                             2                            3
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:09:19,974 - INFO - machamp.model.callback - Performance of 0.7880 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_3.pt
2025-07-10 11:09:21,768 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_2.pt
2025-07-10 11:09:21,912 - INFO - machamp.model.trainer - 
2025-07-10 11:09:21,912 - INFO - machamp.model.trainer - Epoch 4/60: training
2025-07-10 11:09:48,686 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:09:48,686 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 3.909152476780186e-06
2025-07-10 11:09:48,686 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 3.909152476780186e-06
2025-07-10 11:09:48,686 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 1.4854779411764708e-06
2025-07-10 11:09:48,686 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:09:48,686 - INFO - machamp.model.trainer - Epoch 4: evaluating on dev
2025-07-10 11:09:50,362 - INFO - machamp.model.callback - epoch       : 4/60
2025-07-10 11:09:50,362 - INFO - machamp.model.callback - best_epoch  : 4
2025-07-10 11:09:50,362 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:09:50,362 - INFO - machamp.model.callback - cur_ram     : 2.1371
2025-07-10 11:09:50,362 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:09:50,362 - INFO - machamp.model.callback - time_total  : 0:01:41
2025-07-10 11:09:50,362 - INFO - machamp.model.callback - 
2025-07-10 11:09:50,362 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:09:50,362 - INFO - machamp.model.callback - Best (4)                                                   
2025-07-10 11:09:50,362 - INFO - machamp.model.callback - dependency_las     4.2037   3.5534       0.1566     0.2023 
2025-07-10 11:09:50,362 - INFO - machamp.model.callback - lemma_accuracy     4.0776   3.8906       0.0588     0.1063 
2025-07-10 11:09:50,363 - INFO - machamp.model.callback - morph_accuracy     5.7164   5.5917       0.0656     0.0939 
2025-07-10 11:09:50,363 - INFO - machamp.model.callback - upos_accuracy      1.2393   1.1977       0.3605     0.3740 
2025-07-10 11:09:50,363 - INFO - machamp.model.callback - xpos_accuracy      1.5074   1.4123       0.3643     0.5975 
2025-07-10 11:09:50,363 - INFO - machamp.model.callback - sum               16.7444  15.6457       1.0059     1.3740 
2025-07-10 11:09:50,363 - INFO - machamp.model.callback - Epoch 4                                                    
2025-07-10 11:09:50,363 - INFO - machamp.model.callback - dependency_las     4.2037   3.5534       0.1566     0.2023 
2025-07-10 11:09:50,363 - INFO - machamp.model.callback - lemma_accuracy     4.0776   3.8906       0.0588     0.1063 
2025-07-10 11:09:50,363 - INFO - machamp.model.callback - morph_accuracy     5.7164   5.5917       0.0656     0.0939 
2025-07-10 11:09:50,363 - INFO - machamp.model.callback - upos_accuracy      1.2393   1.1977       0.3605     0.3740 
2025-07-10 11:09:50,363 - INFO - machamp.model.callback - xpos_accuracy      1.5074   1.4123       0.3643     0.5975 
2025-07-10 11:09:50,363 - INFO - machamp.model.callback - sum               16.7444  15.6457       1.0059     1.3740 
2025-07-10 11:09:50,371 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                                                          ▗▞│
│                                                        ▗▞▘ │
│                                                       ▄▘   │ 0.5
│                                                     ▄▀     │
│                                                   ▗▞       │
│                                                 ▗▞▘        │ 0.4
│                                  ▗▄▄▄▄▄▄▄▄▄▄▄▄▄▄▘▄▄▄▄▄▄▄▞▀▀│
│                  ▄▄▄▄▄▄▄▄▀▀▀▀▀▀▀▀▘           ▗▀            │
│  ▄▄▄▄▄▄▄▄▀▀▀▀▀▀▀▀                          ▗▞▘             │ 0.3
│▀▀                                         ▞▘               │
│                                         ▄▀                 │
│                                      ▄▄▀▄▄▄▄▄▄▄▄▄▄▄▄▞▀▀▀▀▀▀│ 0.2
│                 ▗▄▄▄▄▄▄▄▄▄▄▄▞▀▀▀▄▄▞▀▀▀▀▘                   │
│▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▘          ▄▄▞▀▀                          ▄│
│                      ▗▄▄▀▀▀                     ▗▄▄▄▄▄▄▞▀▀▀│ 0.1
│   ▗▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▀▀▘▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▞▀▀▀▀▀▀▘▀▘        │
│▀▀▀▘▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▀▀▀▀▀▀▘                │
└────────────────────────────────────────────────────────────┘
 1                   2                  3                   4
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:09:50,371 - INFO - machamp.model.callback - Performance of 1.3740 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_4.pt
2025-07-10 11:09:52,130 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_3.pt
2025-07-10 11:09:52,259 - INFO - machamp.model.trainer - 
2025-07-10 11:09:52,259 - INFO - machamp.model.trainer - Epoch 5/60: training
2025-07-10 11:10:18,978 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:10:18,978 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 5.003869969040248e-06
2025-07-10 11:10:18,978 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 5.003869969040248e-06
2025-07-10 11:10:18,978 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 1.9014705882352943e-06
2025-07-10 11:10:18,978 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:10:18,978 - INFO - machamp.model.trainer - Epoch 5: evaluating on dev
2025-07-10 11:10:20,600 - INFO - machamp.model.callback - epoch       : 5/60
2025-07-10 11:10:20,600 - INFO - machamp.model.callback - best_epoch  : 5
2025-07-10 11:10:20,600 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:10:20,600 - INFO - machamp.model.callback - cur_ram     : 2.1372
2025-07-10 11:10:20,601 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:10:20,601 - INFO - machamp.model.callback - time_total  : 0:02:11
2025-07-10 11:10:20,601 - INFO - machamp.model.callback - 
2025-07-10 11:10:20,601 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:10:20,601 - INFO - machamp.model.callback - Best (5)                                                   
2025-07-10 11:10:20,601 - INFO - machamp.model.callback - dependency_las     3.9548   3.2856       0.1767     0.2120 
2025-07-10 11:10:20,601 - INFO - machamp.model.callback - lemma_accuracy     3.5348   2.9862       0.3423     0.4710 
2025-07-10 11:10:20,601 - INFO - machamp.model.callback - morph_accuracy     5.3453   4.9880       0.1321     0.2395 
2025-07-10 11:10:20,601 - INFO - machamp.model.callback - upos_accuracy      1.1414   1.0742       0.3904     0.3754 
2025-07-10 11:10:20,601 - INFO - machamp.model.callback - xpos_accuracy      1.2357   1.0041       0.6303     0.6444 
2025-07-10 11:10:20,601 - INFO - machamp.model.callback - sum               15.2121  13.3381       1.6717     1.9423 
2025-07-10 11:10:20,602 - INFO - machamp.model.callback - Epoch 5                                                    
2025-07-10 11:10:20,602 - INFO - machamp.model.callback - dependency_las     3.9548   3.2856       0.1767     0.2120 
2025-07-10 11:10:20,602 - INFO - machamp.model.callback - lemma_accuracy     3.5348   2.9862       0.3423     0.4710 
2025-07-10 11:10:20,602 - INFO - machamp.model.callback - morph_accuracy     5.3453   4.9880       0.1321     0.2395 
2025-07-10 11:10:20,602 - INFO - machamp.model.callback - upos_accuracy      1.1414   1.0742       0.3904     0.3754 
2025-07-10 11:10:20,602 - INFO - machamp.model.callback - xpos_accuracy      1.2357   1.0041       0.6303     0.6444 
2025-07-10 11:10:20,602 - INFO - machamp.model.callback - sum               15.2121  13.3381       1.6717     1.9423 
2025-07-10 11:10:20,610 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                                                  ▗▄▄▄▄▄▄▀▀▀│
│                                            ▟▀▀▀▀▀▘         │ 0.6
│                                          ▗▞                │
│                                         ▗▘                 │
│                                        ▞▘                 ▗│
│                                      ▗▀                 ▗▞▘│
│                                     ▄▘                 ▄▘  │ 0.4
│                      ▄▄▄▄▄▄▞▀▀▀▀▀▀▗▞▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀│
│         ▄▄▄▄▄▄▞▀▀▀▀▀▀            ▗▘                 ▞▘     │
│▄▄▞▀▀▀▀▀▀                        ▞▘                ▗▀       │
│                               ▗▀                 ▞▘      ▗▄│
│                              ▄▘     ▗▄▄▄▄▄▄▄▄▄▄▄▀▄▄▄▄▗▄▞▀▘▀│ 0.2
│                   ▄▄▄▄▄▄▄▄▄▀▀▘▀▀▀▀▀▀▘        ▗▞  ▗▄▞▀▘     │
│▄▄▄▄▄▄▀▀▀▀▀▀▀▀▀▀▀▀▀   ▄▄▀▀                   ▄▗▄▞▀▘         │
│                 ▗▄▞▀▀                ▗▄▄▄▄▄▞▀▘             │
│    ▄▄▄▄▄▄▄▄▄▄▄▞▀▘▗▄▄▄▄▄▄▄▄▄▄▄▄▄▄▀▀▀▀▀▘▘                    │
│▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▘▄▄▄▄▄▄▄▄▄▄▞▀▀▀                           │
└────────────────────────────────────────────────────────────┘
 1              2              3             4              5
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:10:20,611 - INFO - machamp.model.callback - Performance of 1.9423 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_5.pt
2025-07-10 11:10:22,360 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_4.pt
2025-07-10 11:10:22,486 - INFO - machamp.model.trainer - 
2025-07-10 11:10:22,486 - INFO - machamp.model.trainer - Epoch 6/60: training
2025-07-10 11:10:49,203 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:10:49,203 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 6.098587461300309e-06
2025-07-10 11:10:49,204 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 6.098587461300309e-06
2025-07-10 11:10:49,204 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 2.3174632352941178e-06
2025-07-10 11:10:49,204 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:10:49,204 - INFO - machamp.model.trainer - Epoch 6: evaluating on dev
2025-07-10 11:10:50,835 - INFO - machamp.model.callback - epoch       : 6/60
2025-07-10 11:10:50,835 - INFO - machamp.model.callback - best_epoch  : 6
2025-07-10 11:10:50,835 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:10:50,835 - INFO - machamp.model.callback - cur_ram     : 2.1398
2025-07-10 11:10:50,835 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:10:50,835 - INFO - machamp.model.callback - time_total  : 0:02:41
2025-07-10 11:10:50,835 - INFO - machamp.model.callback - 
2025-07-10 11:10:50,836 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:10:50,836 - INFO - machamp.model.callback - Best (6)                                                   
2025-07-10 11:10:50,836 - INFO - machamp.model.callback - dependency_las     3.8325   2.9473       0.1857     0.2539 
2025-07-10 11:10:50,836 - INFO - machamp.model.callback - lemma_accuracy     2.5914   2.1994       0.4586     0.4710 
2025-07-10 11:10:50,836 - INFO - machamp.model.callback - morph_accuracy     4.7186   4.3783       0.2348     0.2311 
2025-07-10 11:10:50,836 - INFO - machamp.model.callback - upos_accuracy      1.0426   1.0045       0.3963     0.3848 
2025-07-10 11:10:50,836 - INFO - machamp.model.callback - xpos_accuracy      0.9269   0.8360       0.6476     0.6444 
2025-07-10 11:10:50,836 - INFO - machamp.model.callback - sum               13.1120  11.3654       1.9230     1.9852 
2025-07-10 11:10:50,836 - INFO - machamp.model.callback - Epoch 6                                                    
2025-07-10 11:10:50,836 - INFO - machamp.model.callback - dependency_las     3.8325   2.9473       0.1857     0.2539 
2025-07-10 11:10:50,836 - INFO - machamp.model.callback - lemma_accuracy     2.5914   2.1994       0.4586     0.4710 
2025-07-10 11:10:50,836 - INFO - machamp.model.callback - morph_accuracy     4.7186   4.3783       0.2348     0.2311 
2025-07-10 11:10:50,836 - INFO - machamp.model.callback - upos_accuracy      1.0426   1.0045       0.3963     0.3848 
2025-07-10 11:10:50,836 - INFO - machamp.model.callback - xpos_accuracy      0.9269   0.8360       0.6476     0.6444 
2025-07-10 11:10:50,836 - INFO - machamp.model.callback - sum               13.1120  11.3654       1.9230     1.9852 
2025-07-10 11:10:50,844 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                                        ▗▄▄▄▄▞▀▀▀▀▀▀▀▀▀▀▀▀▀▀│
│                                   ▐▀▀▀▀▘                   │ 0.6
│                                  ▄▘                        │
│                                 ▞                          │
│                                ▞              ▗▄▄▄▄▄▄▄▄▄▄▄▄│
│                               ▞              ▄▘            │
│                              ▞              ▞             ▄│ 0.4
│                  ▄▄▄▄▄▞▀▀▀▀▗▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀ │
│       ▗▄▄▄▄▄▀▀▀▀▀         ▗▘             ▗▘                │
│▄▄▞▀▀▀▀▘                  ▗▘             ▄▘                 │
│                         ▗▘             ▞      ▄▄▄▄▄▄▄▄▄▄▄▄▄│
│                        ▞▘     ▄▄▄▄▄▄▄▗▞▄▄▄▄▄▞▀▀▀▀▀▀▀       │ 0.2
│                ▄▄▄▄▄▄▞▀▘▀▀▀▀▀▀      ▗▘ ▗▄▞▀                │
│▄▄▄▄▄▄▞▀▀▀▀▀▀▀▀▀ ▗▄▞▀               ▄▗▄▀▘                   │
│              ▗▄▀▘             ▗▄▄▄▄▀▘                      │
│     ▄▄▄▄▄▄▄▞▀▘  ▗▄▄▄▄▄▄▄▄▄▀▀▀▀▘                            │
│▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▘▄▄▄▄▄▄▞▀▀                                 │
└────────────────────────────────────────────────────────────┘
 1           2           3          4           5           6
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:10:50,845 - INFO - machamp.model.callback - Performance of 1.9852 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_6.pt
2025-07-10 11:10:52,548 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_5.pt
2025-07-10 11:10:52,678 - INFO - machamp.model.trainer - 
2025-07-10 11:10:52,679 - INFO - machamp.model.trainer - Epoch 7/60: training
2025-07-10 11:11:19,429 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:11:19,429 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 7.1933049535603715e-06
2025-07-10 11:11:19,429 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 7.1933049535603715e-06
2025-07-10 11:11:19,429 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 2.7334558823529413e-06
2025-07-10 11:11:19,429 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:11:19,429 - INFO - machamp.model.trainer - Epoch 7: evaluating on dev
2025-07-10 11:11:21,050 - INFO - machamp.model.callback - epoch       : 7/60
2025-07-10 11:11:21,050 - INFO - machamp.model.callback - best_epoch  : 7
2025-07-10 11:11:21,050 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:11:21,050 - INFO - machamp.model.callback - cur_ram     : 2.1399
2025-07-10 11:11:21,050 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:11:21,050 - INFO - machamp.model.callback - time_total  : 0:03:11
2025-07-10 11:11:21,050 - INFO - machamp.model.callback - 
2025-07-10 11:11:21,050 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:11:21,050 - INFO - machamp.model.callback - Best (7)                                                   
2025-07-10 11:11:21,050 - INFO - machamp.model.callback - dependency_las     3.4506   2.6648       0.2224     0.2929 
2025-07-10 11:11:21,050 - INFO - machamp.model.callback - lemma_accuracy     2.1040   1.9016       0.4585     0.4713 
2025-07-10 11:11:21,050 - INFO - machamp.model.callback - morph_accuracy     4.2516   3.9910       0.2317     0.2295 
2025-07-10 11:11:21,050 - INFO - machamp.model.callback - upos_accuracy      0.9902   0.9524       0.4255     0.4633 
2025-07-10 11:11:21,050 - INFO - machamp.model.callback - xpos_accuracy      0.8409   0.7875       0.6475     0.6444 
2025-07-10 11:11:21,050 - INFO - machamp.model.callback - sum               11.6373  10.2973       1.9856     2.1013 
2025-07-10 11:11:21,050 - INFO - machamp.model.callback - Epoch 7                                                    
2025-07-10 11:11:21,050 - INFO - machamp.model.callback - dependency_las     3.4506   2.6648       0.2224     0.2929 
2025-07-10 11:11:21,050 - INFO - machamp.model.callback - lemma_accuracy     2.1040   1.9016       0.4585     0.4713 
2025-07-10 11:11:21,051 - INFO - machamp.model.callback - morph_accuracy     4.2516   3.9910       0.2317     0.2295 
2025-07-10 11:11:21,051 - INFO - machamp.model.callback - upos_accuracy      0.9902   0.9524       0.4255     0.4633 
2025-07-10 11:11:21,051 - INFO - machamp.model.callback - xpos_accuracy      0.8409   0.7875       0.6475     0.6444 
2025-07-10 11:11:21,051 - INFO - machamp.model.callback - sum               11.6373  10.2973       1.9856     2.1013 
2025-07-10 11:11:21,057 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                                  ▄▄▄▄▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀│
│                             ▗▀▀▀▀                          │ 0.6
│                            ▗▘                              │
│                           ▗▘                               │
│                          ▗▘           ▗▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▗│
│                          ▌           ▗▘              ▗▄▄▀▀▘│
│                         ▞           ▗▘          ▄▄▄▀▀▘     │ 0.4
│               ▄▄▄▄▞▀▀▀▀▞▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀           │
│      ▗▄▄▄▞▀▀▀▀        ▞           ▗▘                       │
│▄▄▀▀▀▀▘               ▞           ▗▘                 ▗▄▄▄▄▞▀│
│                     ▞           ▗▘    ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄ │
│                    ▞     ▄▄▄▄▄▄▗▘▄▄▗▄▀▀▀▀▀▀               ▀│ 0.2
│             ▗▄▄▄▗▄▀▘▀▀▀▀▀     ▄▘ ▄▞▘                       │
│▄▄▄▄▄▞▀▀▀▀▀▀▀▘ ▄▀▘            ▞▗▞▀                          │
│            ▄▞▀           ▗▄▄▄▛▘                            │
│    ▄▄▄▄▄▄▞▀  ▗▄▄▄▄▄▄▄▞▀▀▀▘                                 │
│▀▀▀▀▀▀▀▀▀▀▀▀▀▀▘▄▄▄▄▄▞▀▘                                     │
└────────────────────────────────────────────────────────────┘
 1         2         3         4        5         6         7
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:11:21,058 - INFO - machamp.model.callback - Performance of 2.1013 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_7.pt
2025-07-10 11:11:22,870 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_6.pt
2025-07-10 11:11:23,003 - INFO - machamp.model.trainer - 
2025-07-10 11:11:23,003 - INFO - machamp.model.trainer - Epoch 8/60: training
2025-07-10 11:11:49,763 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:11:49,764 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 8.288022445820435e-06
2025-07-10 11:11:49,764 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 8.288022445820435e-06
2025-07-10 11:11:49,764 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 3.149448529411765e-06
2025-07-10 11:11:49,764 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:11:49,764 - INFO - machamp.model.trainer - Epoch 8: evaluating on dev
2025-07-10 11:11:51,427 - INFO - machamp.model.callback - epoch       : 8/60
2025-07-10 11:11:51,427 - INFO - machamp.model.callback - best_epoch  : 8
2025-07-10 11:11:51,427 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:11:51,427 - INFO - machamp.model.callback - cur_ram     : 2.1270
2025-07-10 11:11:51,427 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:11:51,427 - INFO - machamp.model.callback - time_total  : 0:03:42
2025-07-10 11:11:51,427 - INFO - machamp.model.callback - 
2025-07-10 11:11:51,427 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:11:51,427 - INFO - machamp.model.callback - Best (8)                                                   
2025-07-10 11:11:51,427 - INFO - machamp.model.callback - dependency_las     3.0877   2.4196       0.2582     0.3391 
2025-07-10 11:11:51,427 - INFO - machamp.model.callback - lemma_accuracy     1.8620   1.7213       0.4609     0.4753 
2025-07-10 11:11:51,427 - INFO - machamp.model.callback - morph_accuracy     3.9356   3.7183       0.2365     0.2469 
2025-07-10 11:11:51,427 - INFO - machamp.model.callback - upos_accuracy      0.9396   0.8982       0.4765     0.4867 
2025-07-10 11:11:51,428 - INFO - machamp.model.callback - xpos_accuracy      0.7963   0.7543       0.6479     0.6444 
2025-07-10 11:11:51,428 - INFO - machamp.model.callback - sum               10.6212   9.5117       2.0801     2.1926 
2025-07-10 11:11:51,428 - INFO - machamp.model.callback - Epoch 8                                                    
2025-07-10 11:11:51,428 - INFO - machamp.model.callback - dependency_las     3.0877   2.4196       0.2582     0.3391 
2025-07-10 11:11:51,428 - INFO - machamp.model.callback - lemma_accuracy     1.8620   1.7213       0.4609     0.4753 
2025-07-10 11:11:51,428 - INFO - machamp.model.callback - morph_accuracy     3.9356   3.7183       0.2365     0.2469 
2025-07-10 11:11:51,428 - INFO - machamp.model.callback - upos_accuracy      0.9396   0.8982       0.4765     0.4867 
2025-07-10 11:11:51,428 - INFO - machamp.model.callback - xpos_accuracy      0.7963   0.7543       0.6479     0.6444 
2025-07-10 11:11:51,428 - INFO - machamp.model.callback - sum               10.6212   9.5117       2.0801     2.1926 
2025-07-10 11:11:51,435 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                             ▄▄▄▞▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀│
│                         ▐▀▀▀                               │ 0.6
│                        ▗▘                                  │
│                       ▗▘                                   │
│                       ▞          ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▞▀▀│
│                      ▞          ▞             ▄▄▀▀         │
│                     ▐          ▞         ▄▄▞▀▀             │ 0.4
│             ▄▄▄▞▀▀▀▗▘▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀                  │
│     ▗▄▄▄▀▀▀▀      ▗▘         ▗▘                      ▄▄▄▞▀▀│
│▄▄▀▀▀▘             ▞         ▗▘               ▄▄▄▄▀▀▀▀      │
│                  ▞         ▗▘   ▗▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▖▗▄▄▄▄▄▄▄▄│
│                 ▐    ▗▄▄▄▄▗▘▄▄▗▞▘▀▀▀▀            ▝▘        │ 0.2
│           ▗▄▄▄▄▞▘▀▀▀▀▘    ▞ ▄▞▘                            │
│▄▄▄▄▞▀▀▀▀▀▀▘▗▄▀           ▞▄▀                               │
│          ▗▞▘         ▗▄▄▄▀                                 │
│   ▗▄▄▄▄▄▀▘ ▗▄▄▄▄▄▄▞▀▀▘                                     │
│▀▀▀▘▀▀▀▀▀▀▀▀▘▄▄▄▄▞▀▘                                        │
└────────────────────────────────────────────────────────────┘
         2                4                6                8
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:11:51,435 - INFO - machamp.model.callback - Performance of 2.1926 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_8.pt
2025-07-10 11:11:53,177 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_7.pt
2025-07-10 11:11:53,306 - INFO - machamp.model.trainer - 
2025-07-10 11:11:53,306 - INFO - machamp.model.trainer - Epoch 9/60: training
2025-07-10 11:12:20,035 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:12:20,035 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 9.382739938080496e-06
2025-07-10 11:12:20,035 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 9.382739938080496e-06
2025-07-10 11:12:20,035 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 3.5654411764705887e-06
2025-07-10 11:12:20,035 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:12:20,036 - INFO - machamp.model.trainer - Epoch 9: evaluating on dev
2025-07-10 11:12:21,676 - INFO - machamp.model.callback - epoch       : 9/60
2025-07-10 11:12:21,676 - INFO - machamp.model.callback - best_epoch  : 9
2025-07-10 11:12:21,677 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:12:21,677 - INFO - machamp.model.callback - cur_ram     : 2.1354
2025-07-10 11:12:21,677 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:12:21,677 - INFO - machamp.model.callback - time_total  : 0:04:12
2025-07-10 11:12:21,677 - INFO - machamp.model.callback - 
2025-07-10 11:12:21,677 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:12:21,677 - INFO - machamp.model.callback - Best (9)                                                   
2025-07-10 11:12:21,677 - INFO - machamp.model.callback - dependency_las     2.8192   2.2308       0.2942     0.3824 
2025-07-10 11:12:21,677 - INFO - machamp.model.callback - lemma_accuracy     1.7035   1.5809       0.4796     0.5156 
2025-07-10 11:12:21,677 - INFO - machamp.model.callback - morph_accuracy     3.7103   3.5048       0.2600     0.3080 
2025-07-10 11:12:21,677 - INFO - machamp.model.callback - upos_accuracy      0.8837   0.8346       0.5044     0.5129 
2025-07-10 11:12:21,677 - INFO - machamp.model.callback - xpos_accuracy      0.7653   0.7203       0.6476     0.6444 
2025-07-10 11:12:21,677 - INFO - machamp.model.callback - sum                9.8820   8.8715       2.1858     2.3633 
2025-07-10 11:12:21,677 - INFO - machamp.model.callback - Epoch 9                                                    
2025-07-10 11:12:21,677 - INFO - machamp.model.callback - dependency_las     2.8192   2.2308       0.2942     0.3824 
2025-07-10 11:12:21,677 - INFO - machamp.model.callback - lemma_accuracy     1.7035   1.5809       0.4796     0.5156 
2025-07-10 11:12:21,677 - INFO - machamp.model.callback - morph_accuracy     3.7103   3.5048       0.2600     0.3080 
2025-07-10 11:12:21,677 - INFO - machamp.model.callback - upos_accuracy      0.8837   0.8346       0.5044     0.5129 
2025-07-10 11:12:21,677 - INFO - machamp.model.callback - xpos_accuracy      0.7653   0.7203       0.6476     0.6444 
2025-07-10 11:12:21,677 - INFO - machamp.model.callback - sum                9.8820   8.8715       2.1858     2.3633 
2025-07-10 11:12:21,684 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                         ▗▄▄▞▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀│
│                      ▞▀▀▘                                  │ 0.6
│                     ▗▘                                     │
│                    ▗▘                                  ▄▄▄▄│
│                    ▞         ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▗▄▄▄▄▄▞▀▀▀▀▀▀   │
│                   ▐         ▞           ▄▄▀▘               │
│                  ▗▘        ▞        ▄▄▀▀                   │ 0.4
│           ▗▄▄▞▀▀▀▞▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀                ▗▄▄▞▀▀▀│
│     ▄▄▄▀▀▀▘     ▞        ▗▘                   ▄▄▄▀▀▀▘     ▗│
│▄▞▀▀▀           ▗▘        ▌             ▄▄▄▄▀▀▀        ▄▄▀▀▘│
│                ▌        ▞   ▗▄▄▄▄▄▄▄▄▄▄▄▄▄▄ ▄▄▄▄▄▄▄▞▀▀     │
│               ▞   ▗▄▄▄▄▞▄▄▗▞▘▀▀▀           ▀               │ 0.2
│          ▄▄▄▗▞▘▀▀▀▘   ▗▘▗▞▘                                │
│▄▄▄▄▀▀▀▀▀▀ ▄▀▘        ▗▗▞▘                                  │
│         ▄▀         ▄▄▟▘                                    │
│   ▄▄▄▄▄▀  ▄▄▄▄▄▄▀▀▀                                        │
│▀▀▀▀▀▀▀▀▀▀▀▄▄▄▄▞▀                                           │
└────────────────────────────────────────────────────────────┘
        2              4              6              8
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:12:21,684 - INFO - machamp.model.callback - Performance of 2.3633 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_9.pt
2025-07-10 11:12:23,637 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_8.pt
2025-07-10 11:12:23,773 - INFO - machamp.model.trainer - 
2025-07-10 11:12:23,774 - INFO - machamp.model.trainer - Epoch 10/60: training
2025-07-10 11:12:50,475 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:12:50,475 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.0477457430340556e-05
2025-07-10 11:12:50,475 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.0477457430340556e-05
2025-07-10 11:12:50,475 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 3.981433823529412e-06
2025-07-10 11:12:50,475 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:12:50,475 - INFO - machamp.model.trainer - Epoch 10: evaluating on dev
2025-07-10 11:12:52,093 - INFO - machamp.model.callback - epoch       : 10/60
2025-07-10 11:12:52,094 - INFO - machamp.model.callback - best_epoch  : 10
2025-07-10 11:12:52,094 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:12:52,094 - INFO - machamp.model.callback - cur_ram     : 2.1354
2025-07-10 11:12:52,094 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:12:52,094 - INFO - machamp.model.callback - time_total  : 0:04:42
2025-07-10 11:12:52,094 - INFO - machamp.model.callback - 
2025-07-10 11:12:52,094 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:12:52,094 - INFO - machamp.model.callback - Best (10)                                                  
2025-07-10 11:12:52,094 - INFO - machamp.model.callback - dependency_las     2.5513   2.0636       0.3382     0.4280 
2025-07-10 11:12:52,094 - INFO - machamp.model.callback - lemma_accuracy     1.5733   1.4640       0.5164     0.5609 
2025-07-10 11:12:52,094 - INFO - machamp.model.callback - morph_accuracy     3.5140   3.3073       0.3125     0.3703 
2025-07-10 11:12:52,094 - INFO - machamp.model.callback - upos_accuracy      0.8176   0.7643       0.5316     0.5555 
2025-07-10 11:12:52,094 - INFO - machamp.model.callback - xpos_accuracy      0.7340   0.6859       0.6474     0.6444 
2025-07-10 11:12:52,094 - INFO - machamp.model.callback - sum                9.1902   8.2851       2.3461     2.5592 
2025-07-10 11:12:52,094 - INFO - machamp.model.callback - Epoch 10                                                   
2025-07-10 11:12:52,094 - INFO - machamp.model.callback - dependency_las     2.5513   2.0636       0.3382     0.4280 
2025-07-10 11:12:52,094 - INFO - machamp.model.callback - lemma_accuracy     1.5733   1.4640       0.5164     0.5609 
2025-07-10 11:12:52,094 - INFO - machamp.model.callback - morph_accuracy     3.5140   3.3073       0.3125     0.3703 
2025-07-10 11:12:52,095 - INFO - machamp.model.callback - upos_accuracy      0.8176   0.7643       0.5316     0.5555 
2025-07-10 11:12:52,095 - INFO - machamp.model.callback - xpos_accuracy      0.7340   0.6859       0.6474     0.6444 
2025-07-10 11:12:52,095 - INFO - machamp.model.callback - sum                9.1902   8.2851       2.3461     2.5592 
2025-07-10 11:12:52,102 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                      ▗▄▄▞▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀│
│                   ▗▀▀▘                                     │ 0.6
│                   ▞                                    ▗▄▄▄│
│                  ▗▘                              ▄▄▄▄▀▀▀   │
│                  ▌       ▗▄▄▄▄▄▄▄▄▄▄▄▄▗▄▄▄▄▄▀▀▀▀▀▘         │
│                 ▞        ▌         ▗▄▀▘                   ▄│
│                ▗▘       ▞       ▄▄▀▘                ▗▄▄▀▀▀ │ 0.4
│          ▄▄▄▀▀▀▞▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀              ▗▄▄▞▀▀▘   ▄▄▀│
│    ▗▄▄▀▀▀     ▐       ▗▘                 ▄▄▞▀▀▘     ▄▄▀▀   │
│▄▞▀▀▘         ▗▘       ▞           ▗▄▄▄▀▀▀       ▄▄▀▀       │
│              ▞       ▐   ▗▄▄▄▄▄▄▄▄▄▄▄▄▖▄▄▄▄▄▄▞▀▀           │
│             ▗▘  ▗▄▄▄▗▘▄▗▞▘▀▀▘         ▝                    │ 0.2
│         ▄▄▄▞▀▀▀▀▘   ▌▗▞▘                                   │
│▄▄▄▞▀▀▀▀▀ ▄▀        ▞▄▘                                     │
│        ▄▀       ▗▄▄▛                                       │
│   ▄▄▄▄▀ ▗▄▄▄▄▄▀▀▘                                          │
│▀▀▀▀▀▀▀▀▀▘▄▄▄▟▀                                             │
└────────────────────────────────────────────────────────────┘
       2             4            6            8           10
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:12:52,102 - INFO - machamp.model.callback - Performance of 2.5592 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_10.pt
2025-07-10 11:12:53,894 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_9.pt
2025-07-10 11:12:54,023 - INFO - machamp.model.trainer - 
2025-07-10 11:12:54,023 - INFO - machamp.model.trainer - Epoch 11/60: training
2025-07-10 11:13:20,846 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:13:20,846 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.1572174922600619e-05
2025-07-10 11:13:20,846 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.1572174922600619e-05
2025-07-10 11:13:20,846 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 4.397426470588235e-06
2025-07-10 11:13:20,846 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:13:20,846 - INFO - machamp.model.trainer - Epoch 11: evaluating on dev
2025-07-10 11:13:22,489 - INFO - machamp.model.callback - epoch       : 11/60
2025-07-10 11:13:22,489 - INFO - machamp.model.callback - best_epoch  : 11
2025-07-10 11:13:22,489 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:13:22,489 - INFO - machamp.model.callback - cur_ram     : 2.1378
2025-07-10 11:13:22,489 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:13:22,489 - INFO - machamp.model.callback - time_total  : 0:05:13
2025-07-10 11:13:22,489 - INFO - machamp.model.callback - 
2025-07-10 11:13:22,489 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:13:22,489 - INFO - machamp.model.callback - Best (11)                                                  
2025-07-10 11:13:22,489 - INFO - machamp.model.callback - dependency_las     2.3683   1.9158       0.3785     0.4535 
2025-07-10 11:13:22,489 - INFO - machamp.model.callback - lemma_accuracy     1.4590   1.3538       0.5464     0.5827 
2025-07-10 11:13:22,489 - INFO - machamp.model.callback - morph_accuracy     3.3146   3.0893       0.3610     0.3911 
2025-07-10 11:13:22,490 - INFO - machamp.model.callback - upos_accuracy      0.7467   0.6870       0.5730     0.6045 
2025-07-10 11:13:22,490 - INFO - machamp.model.callback - xpos_accuracy      0.6939   0.6424       0.6500     0.6511 
2025-07-10 11:13:22,490 - INFO - machamp.model.callback - sum                8.5825   7.6884       2.5089     2.6830 
2025-07-10 11:13:22,490 - INFO - machamp.model.callback - Epoch 11                                                   
2025-07-10 11:13:22,490 - INFO - machamp.model.callback - dependency_las     2.3683   1.9158       0.3785     0.4535 
2025-07-10 11:13:22,490 - INFO - machamp.model.callback - lemma_accuracy     1.4590   1.3538       0.5464     0.5827 
2025-07-10 11:13:22,490 - INFO - machamp.model.callback - morph_accuracy     3.3146   3.0893       0.3610     0.3911 
2025-07-10 11:13:22,490 - INFO - machamp.model.callback - upos_accuracy      0.7467   0.6870       0.5730     0.6045 
2025-07-10 11:13:22,490 - INFO - machamp.model.callback - xpos_accuracy      0.6939   0.6424       0.6500     0.6511 
2025-07-10 11:13:22,490 - INFO - machamp.model.callback - sum                8.5825   7.6884       2.5089     2.6830 
2025-07-10 11:13:22,498 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                     ▗▄▄▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀│
│                  ▄▀▀▘                                   ▄▄▞│ 0.6
│                 ▞                                 ▗▄▄▄▀▀▀▀▀│
│                ▗▘                            ▄▄▄▞▀▀        │
│                ▞       ▄▄▄▄▄▄▄▄▄▄▄▄▗▄▄▄▄▞▀▀▀▀▘             │
│               ▗▘      ▐         ▄▞▀▘                ▗▄▄▄▄▀▀│
│               ▞      ▗▘      ▗▞▀               ▗▄▄▀▀▘     ▄│ 0.4
│         ▗▄▄▄▀▐▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▘            ▄▄▄▀▀▘  ▗▄▞▀▀▀▀▀ │
│    ▗▄▄▀▀▘    ▌      ▐                ▄▄▞▀▀     ▄▄▀▘        │
│▄▄▀▀▘        ▐      ▗▘          ▗▄▄▞▀▀      ▄▄▀▀            │
│            ▗▘      ▞   ▄▄▄▄▄▖▀▀▘    ▄▄▄▄▄▞▀                │
│            ▞   ▄▄▄▗▘▄▄▀▛▀▀  ▝▀▀▀▀▀▀▀                       │ 0.2
│        ▗▄▗▞▘▀▀▀   ▌▗▞                                      │
│▄▄▄▞▀▀▀▀▘▄▘       ▞▄▘                                       │
│       ▗▀       ▄▄▛                                         │
│  ▗▄▄▄▞▘ ▄▄▄▄▞▀▀                                            │
│▀▀▘▀▀▀▀▀▀▄▄▄▞▘                                              │
└────────────────────────────────────────────────────────────┘
       2           4           6          8          10
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:13:22,498 - INFO - machamp.model.callback - Performance of 2.6830 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_11.pt
2025-07-10 11:13:24,481 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_10.pt
2025-07-10 11:13:24,609 - INFO - machamp.model.trainer - 
2025-07-10 11:13:24,610 - INFO - machamp.model.trainer - Epoch 12/60: training
2025-07-10 11:13:51,248 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:13:51,248 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.2666892414860682e-05
2025-07-10 11:13:51,248 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.2666892414860682e-05
2025-07-10 11:13:51,248 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 4.813419117647059e-06
2025-07-10 11:13:51,248 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:13:51,248 - INFO - machamp.model.trainer - Epoch 12: evaluating on dev
2025-07-10 11:13:52,902 - INFO - machamp.model.callback - epoch       : 12/60
2025-07-10 11:13:52,902 - INFO - machamp.model.callback - best_epoch  : 12
2025-07-10 11:13:52,902 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:13:52,902 - INFO - machamp.model.callback - cur_ram     : 2.1424
2025-07-10 11:13:52,902 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:13:52,902 - INFO - machamp.model.callback - time_total  : 0:05:43
2025-07-10 11:13:52,903 - INFO - machamp.model.callback - 
2025-07-10 11:13:52,903 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:13:52,903 - INFO - machamp.model.callback - Best (12)                                                  
2025-07-10 11:13:52,903 - INFO - machamp.model.callback - dependency_las     2.1678   1.8906       0.4102     0.4764 
2025-07-10 11:13:52,903 - INFO - machamp.model.callback - lemma_accuracy     1.3607   1.2644       0.5797     0.6078 
2025-07-10 11:13:52,903 - INFO - machamp.model.callback - morph_accuracy     3.1136   2.8780       0.4077     0.4515 
2025-07-10 11:13:52,903 - INFO - machamp.model.callback - upos_accuracy      0.6739   0.6065       0.6270     0.6635 
2025-07-10 11:13:52,903 - INFO - machamp.model.callback - xpos_accuracy      0.6552   0.5973       0.6636     0.6645 
2025-07-10 11:13:52,903 - INFO - machamp.model.callback - sum                7.9712   7.2368       2.6882     2.8638 
2025-07-10 11:13:52,903 - INFO - machamp.model.callback - Epoch 12                                                   
2025-07-10 11:13:52,903 - INFO - machamp.model.callback - dependency_las     2.1678   1.8906       0.4102     0.4764 
2025-07-10 11:13:52,903 - INFO - machamp.model.callback - lemma_accuracy     1.3607   1.2644       0.5797     0.6078 
2025-07-10 11:13:52,903 - INFO - machamp.model.callback - morph_accuracy     3.1136   2.8780       0.4077     0.4515 
2025-07-10 11:13:52,903 - INFO - machamp.model.callback - upos_accuracy      0.6739   0.6065       0.6270     0.6635 
2025-07-10 11:13:52,903 - INFO - machamp.model.callback - xpos_accuracy      0.6552   0.5973       0.6636     0.6645 
2025-07-10 11:13:52,903 - INFO - machamp.model.callback - sum                7.9712   7.2368       2.6882     2.8638 
2025-07-10 11:13:52,911 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                     ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▀▀▀▀│
│                ▗▄▞▀▀                                ▗▄▞▀▄▄▄│ 0.6
│                ▌                               ▗▄▄▀▀▘▀▀▀   │
│               ▐                           ▗▄▄▞▀▀           │
│               ▌                   ▗▄▄▄▄▀▀▀▀               ▗│
│              ▐      ▐▀▀▀▀▀▀▀▀▀▄▀▀▀▘▀▀▘           ▗▄▄▄▄▀▀▀▗▄│
│             ▗▘     ▗▘      ▄▞▀              ▄▄▞▀▀▘   ▗▄▞▀▘ │ 0.4
│         ▗▄▄▄▞▄▄▄▄▄▄▄▄▞▀▀▀▀▀            ▄▄▞▀▀  ▗▄▄▞▀▀▀▘     │
│     ▄▄▀▀▘  ▗▘     ▗▘              ▗▄▄▀▀    ▄▄▀▘            │
│▄▄▞▀▀       ▞      ▌          ▗▄▄▀▀▘    ▗▄▞▀                │
│           ▗▘     ▐  ▗▄▄ ▄▄▞▀▀▘    ▄▄▄▄▀▘                   │
│           ▞   ▗▄▗▘▄▗▀▞▀▀▀▀▀▀▀▀▀▀▀▀                         │
│        ▄▄▞▘▀▀▀▘ ▞ ▞▘                                       │ 0.2
│▄▄▄▄▀▀▀▀▗▀      ▗▗▀                                         │
│       ▞▘     ▗▄▟▘                                          │
│  ▗▄▄▄▀ ▗▄▄▄▞▀▘                                             │
│▀▀▘▀▀▀▀▀▘▄▄▞▘                                               │
└────────────────────────────────────────────────────────────┘
         2            5             7            10
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:13:52,911 - INFO - machamp.model.callback - Performance of 2.8638 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_12.pt
2025-07-10 11:13:54,680 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_11.pt
2025-07-10 11:13:54,808 - INFO - machamp.model.trainer - 
2025-07-10 11:13:54,809 - INFO - machamp.model.trainer - Epoch 13/60: training
2025-07-10 11:14:21,436 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:14:21,436 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.3761609907120743e-05
2025-07-10 11:14:21,436 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.3761609907120743e-05
2025-07-10 11:14:21,436 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 5.229411764705883e-06
2025-07-10 11:14:21,436 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:14:21,436 - INFO - machamp.model.trainer - Epoch 13: evaluating on dev
2025-07-10 11:14:23,101 - INFO - machamp.model.callback - epoch       : 13/60
2025-07-10 11:14:23,101 - INFO - machamp.model.callback - best_epoch  : 13
2025-07-10 11:14:23,101 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:14:23,101 - INFO - machamp.model.callback - cur_ram     : 2.1424
2025-07-10 11:14:23,101 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:14:23,101 - INFO - machamp.model.callback - time_total  : 0:06:13
2025-07-10 11:14:23,101 - INFO - machamp.model.callback - 
2025-07-10 11:14:23,101 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:14:23,101 - INFO - machamp.model.callback - Best (13)                                                  
2025-07-10 11:14:23,102 - INFO - machamp.model.callback - dependency_las     2.0088   1.7826       0.4551     0.5039 
2025-07-10 11:14:23,102 - INFO - machamp.model.callback - lemma_accuracy     1.2667   1.1882       0.6054     0.6270 
2025-07-10 11:14:23,102 - INFO - machamp.model.callback - morph_accuracy     2.9027   2.6857       0.4502     0.4676 
2025-07-10 11:14:23,102 - INFO - machamp.model.callback - upos_accuracy      0.6003   0.5396       0.6750     0.6890 
2025-07-10 11:14:23,102 - INFO - machamp.model.callback - xpos_accuracy      0.6081   0.5533       0.6866     0.7038 
2025-07-10 11:14:23,102 - INFO - machamp.model.callback - sum                7.3867   6.7493       2.8723     2.9913 
2025-07-10 11:14:23,102 - INFO - machamp.model.callback - Epoch 13                                                   
2025-07-10 11:14:23,102 - INFO - machamp.model.callback - dependency_las     2.0088   1.7826       0.4551     0.5039 
2025-07-10 11:14:23,102 - INFO - machamp.model.callback - lemma_accuracy     1.2667   1.1882       0.6054     0.6270 
2025-07-10 11:14:23,102 - INFO - machamp.model.callback - morph_accuracy     2.9027   2.6857       0.4502     0.4676 
2025-07-10 11:14:23,102 - INFO - machamp.model.callback - upos_accuracy      0.6003   0.5396       0.6750     0.6890 
2025-07-10 11:14:23,102 - INFO - machamp.model.callback - xpos_accuracy      0.6081   0.5533       0.6866     0.7038 
2025-07-10 11:14:23,102 - INFO - machamp.model.callback - sum                7.3867   6.7493       2.8723     2.9913 
2025-07-10 11:14:23,111 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                                                        ▄▄▞▀│
│                  ▗▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▀▀▀▀▀▀▘   │
│               ▄▞▀▘                             ▄▄▀▀▄▄▄▀▀▀▀▀│ 0.6
│              ▐                            ▗▄▄▀▀▀▀▀         │
│              ▌                       ▄▄▄▞▀▘               ▗│
│             ▐      ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▀▀▀▀▘            ▗▄▄▄▄▀▀▄▄│
│             ▌     ▐       ▄▞▀              ▄▄▄▀▀▀▀▘▄▄▀▀▀▀  │
│            ▐     ▗▘ ▄▄▄▄▄▀            ▗▄▞▀▀  ▗▄▄▄▞▀        │ 0.4
│      ▗▄▄▀▀▀▌▀▀▀▀▀▀▀▀             ▗▄▄▀▀▘  ▄▞▀▀▘             │
│  ▄▄▀▀▘    ▐     ▗▘           ▄▄▀▀▘   ▗▄▞▀                  │
│▀▀         ▌     ▞      ▗▄▄▞▀▀     ▄▄▀▘                     │
│          ▐     ▐ ▄▄▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀                         │
│         ▗▘▄▞▀▀▀▌▗▞                                         │ 0.2
│▗▄▄▄▄▞▀▀▞▘     ▐▄▘                                          │
│▘     ▗▀      ▄▞                                            │
│   ▄▄▞▘ ▗▄▄▄▀▀                                              │
│▀▀▀▀▀▀▀▀▘▄▞▘                                                │
└────────────────────────────────────────────────────────────┘
      2         4         6        8        10        12
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:14:23,111 - INFO - machamp.model.callback - Performance of 2.9913 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_13.pt
2025-07-10 11:14:24,859 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_12.pt
2025-07-10 11:14:24,988 - INFO - machamp.model.trainer - 
2025-07-10 11:14:24,988 - INFO - machamp.model.trainer - Epoch 14/60: training
2025-07-10 11:14:51,691 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:14:51,691 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.4856327399380808e-05
2025-07-10 11:14:51,691 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.4856327399380808e-05
2025-07-10 11:14:51,691 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 5.6454044117647075e-06
2025-07-10 11:14:51,691 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:14:51,692 - INFO - machamp.model.trainer - Epoch 14: evaluating on dev
2025-07-10 11:14:53,318 - INFO - machamp.model.callback - epoch       : 14/60
2025-07-10 11:14:53,318 - INFO - machamp.model.callback - best_epoch  : 14
2025-07-10 11:14:53,318 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:14:53,318 - INFO - machamp.model.callback - cur_ram     : 2.1424
2025-07-10 11:14:53,318 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:14:53,318 - INFO - machamp.model.callback - time_total  : 0:06:43
2025-07-10 11:14:53,318 - INFO - machamp.model.callback - 
2025-07-10 11:14:53,318 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:14:53,318 - INFO - machamp.model.callback - Best (14)                                                  
2025-07-10 11:14:53,318 - INFO - machamp.model.callback - dependency_las     1.8324   1.7193       0.4892     0.5196 
2025-07-10 11:14:53,319 - INFO - machamp.model.callback - lemma_accuracy     1.1860   1.1234       0.6302     0.6525 
2025-07-10 11:14:53,319 - INFO - machamp.model.callback - morph_accuracy     2.7129   2.5051       0.4771     0.5005 
2025-07-10 11:14:53,319 - INFO - machamp.model.callback - upos_accuracy      0.5334   0.4797       0.7020     0.7085 
2025-07-10 11:14:53,319 - INFO - machamp.model.callback - xpos_accuracy      0.5650   0.5117       0.7198     0.7340 
2025-07-10 11:14:53,319 - INFO - machamp.model.callback - sum                6.8296   6.3392       3.0182     3.1151 
2025-07-10 11:14:53,319 - INFO - machamp.model.callback - Epoch 14                                                   
2025-07-10 11:14:53,319 - INFO - machamp.model.callback - dependency_las     1.8324   1.7193       0.4892     0.5196 
2025-07-10 11:14:53,319 - INFO - machamp.model.callback - lemma_accuracy     1.1860   1.1234       0.6302     0.6525 
2025-07-10 11:14:53,319 - INFO - machamp.model.callback - morph_accuracy     2.7129   2.5051       0.4771     0.5005 
2025-07-10 11:14:53,319 - INFO - machamp.model.callback - upos_accuracy      0.5334   0.4797       0.7020     0.7085 
2025-07-10 11:14:53,319 - INFO - machamp.model.callback - xpos_accuracy      0.5650   0.5117       0.7198     0.7340 
2025-07-10 11:14:53,319 - INFO - machamp.model.callback - sum                6.8296   6.3392       3.0182     3.1151 
2025-07-10 11:14:53,327 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                                                       ▄▄▄▀▀│
│                                                ▄▄▄▄▞▀▀▀▀▀  │
│               ▗▄▞▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▞▀  ▄▄▄▄▞▀▀▀│
│             ▗▀▘                          ▗▄▄▀▘▞▀▀▀▀        │ 0.6
│             ▞                        ▄▄▞▀▘                 │
│            ▗▘                 ▄▄▄▄▀▀▀             ▗▄▄▄▀▀▄▄▄│
│            ▞     ▞▀▀▀▀▀▀▀▄▞▀▀▀▀▀          ▄▄▄▄▀▀▗▄▄▞▀▀▀▀   │
│           ▗▘    ▗▘    ▗▄▀            ▄▄▀▀▀   ▄▄▀▘          │ 0.4
│       ▗▄▄▄▞▀▀▀▀▀▀▀▀▀▀▀▘         ▗▄▞▀▀  ▄▞▀▀▀▀              │
│   ▄▄▀▀▘  ▗▘    ▐            ▄▄▀▀▘  ▗▄▀▀                    │
│▞▀▀       ▞     ▌       ▄▄▞▀▀    ▗▄▀▘                       │
│         ▗▘    ▐  ▞▄▄▄▄▄▄▄▄▄▄▄▀▀▀▘                          │
│         ▞▄▄▄▀▗▘▗▀▀                                         │ 0.2
│ ▗▄▄▄▄▀▗▀     ▞▄▘                                           │
│▀▘    ▄▘     ▄▞                                             │
│   ▄▄▞  ▗▄▄▀▀                                               │
│▀▀▀▀▀▀▀▀▘▞▘                                                 │
└────────────────────────────────────────────────────────────┘
       2           5           7         10          12
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:14:53,327 - INFO - machamp.model.callback - Performance of 3.1151 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_14.pt
2025-07-10 11:14:55,146 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_13.pt
2025-07-10 11:14:55,274 - INFO - machamp.model.trainer - 
2025-07-10 11:14:55,275 - INFO - machamp.model.trainer - Epoch 15/60: training
2025-07-10 11:15:21,934 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:15:21,934 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.5951044891640867e-05
2025-07-10 11:15:21,934 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.5951044891640867e-05
2025-07-10 11:15:21,934 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 6.06139705882353e-06
2025-07-10 11:15:21,934 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:15:21,935 - INFO - machamp.model.trainer - Epoch 15: evaluating on dev
2025-07-10 11:15:23,578 - INFO - machamp.model.callback - epoch       : 15/60
2025-07-10 11:15:23,578 - INFO - machamp.model.callback - best_epoch  : 15
2025-07-10 11:15:23,578 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:15:23,578 - INFO - machamp.model.callback - cur_ram     : 2.1276
2025-07-10 11:15:23,578 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:15:23,579 - INFO - machamp.model.callback - time_total  : 0:07:14
2025-07-10 11:15:23,579 - INFO - machamp.model.callback - 
2025-07-10 11:15:23,579 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:15:23,579 - INFO - machamp.model.callback - Best (15)                                                  
2025-07-10 11:15:23,579 - INFO - machamp.model.callback - dependency_las     1.7198   1.7474       0.5105     0.5391 
2025-07-10 11:15:23,579 - INFO - machamp.model.callback - lemma_accuracy     1.1173   1.0634       0.6544     0.6723 
2025-07-10 11:15:23,579 - INFO - machamp.model.callback - morph_accuracy     2.5280   2.3353       0.5113     0.5283 
2025-07-10 11:15:23,579 - INFO - machamp.model.callback - upos_accuracy      0.4751   0.4203       0.7242     0.7457 
2025-07-10 11:15:23,579 - INFO - machamp.model.callback - xpos_accuracy      0.5205   0.4692       0.7420     0.7474 
2025-07-10 11:15:23,579 - INFO - machamp.model.callback - sum                6.3606   6.0355       3.1424     3.2328 
2025-07-10 11:15:23,579 - INFO - machamp.model.callback - Epoch 15                                                   
2025-07-10 11:15:23,579 - INFO - machamp.model.callback - dependency_las     1.7198   1.7474       0.5105     0.5391 
2025-07-10 11:15:23,579 - INFO - machamp.model.callback - lemma_accuracy     1.1173   1.0634       0.6544     0.6723 
2025-07-10 11:15:23,579 - INFO - machamp.model.callback - morph_accuracy     2.5280   2.3353       0.5113     0.5283 
2025-07-10 11:15:23,579 - INFO - machamp.model.callback - upos_accuracy      0.4751   0.4203       0.7242     0.7457 
2025-07-10 11:15:23,579 - INFO - machamp.model.callback - xpos_accuracy      0.5205   0.4692       0.7420     0.7474 
2025-07-10 11:15:23,579 - INFO - machamp.model.callback - sum                6.3606   6.0355       3.1424     3.2328 
2025-07-10 11:15:23,587 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                                                    ▗▄▄▄▞▀▀▀│
│                                               ▗▄▄▀▀▘▀▀▀   ▄│
│               ▗▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▀▀▀▀▀▀▀▘  ▗▄▄▄▞▀▀▀▀ │
│            ▗▞▀▘                        ▗▄▞▀▄▄▀▀▀▀▘         │ 0.6
│            ▐                       ▗▄▞▀▘▀                 ▄│
│            ▌                 ▗▄▄▄▀▀▘            ▄▄▄▄▞▄▄▄▞▀▀│
│           ▐     ▛▀▀▀▀▀▀▀▄▞▀▀▀▘▘         ▗▄▄▄▞▗▄▄▄▄▀▀▀      │
│           ▌    ▐     ▗▞▀           ▄▄▀▀▀▘ ▗▄▀▘             │ 0.4
│       ▗▄▄▐▄▄▄▄▄▄▄▞▀▀▀▘        ▗▄▄▀▀ ▗▄▄▀▀▀▘                │
│   ▗▄▞▀▘  ▌    ▐           ▗▄▞▀▘  ▗▄▀▘                      │
│▄▀▀▘     ▐     ▌       ▄▄▀▀▘   ▗▄▀▘                         │
│         ▌    ▞ ▗▄▄▄▄▄▄▄▄▄▄▄▄▀▀▘                            │
│        ▗▘▄▄▞▗▘▗▘▀                                          │ 0.2
│  ▄▄▄▄▀▞▘    ▞▄▘                                            │
│▀▀   ▗▞     ▗▞                                              │
│   ▄▄▘  ▄▄▞▀▘                                               │
│▀▀▀▀▀▀▀▀▄▀                                                  │
└────────────────────────────────────────────────────────────┘
                  5                   10                   15
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:15:23,587 - INFO - machamp.model.callback - Performance of 3.2328 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_15.pt
2025-07-10 11:15:25,340 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_14.pt
2025-07-10 11:15:25,468 - INFO - machamp.model.trainer - 
2025-07-10 11:15:25,468 - INFO - machamp.model.trainer - Epoch 16/60: training
2025-07-10 11:15:52,053 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:15:52,053 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.704576238390093e-05
2025-07-10 11:15:52,053 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.704576238390093e-05
2025-07-10 11:15:52,053 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 6.477389705882354e-06
2025-07-10 11:15:52,053 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:15:52,053 - INFO - machamp.model.trainer - Epoch 16: evaluating on dev
2025-07-10 11:15:53,666 - INFO - machamp.model.callback - epoch       : 16/60
2025-07-10 11:15:53,666 - INFO - machamp.model.callback - best_epoch  : 16
2025-07-10 11:15:53,666 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:15:53,666 - INFO - machamp.model.callback - cur_ram     : 2.1406
2025-07-10 11:15:53,666 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:15:53,666 - INFO - machamp.model.callback - time_total  : 0:07:44
2025-07-10 11:15:53,666 - INFO - machamp.model.callback - 
2025-07-10 11:15:53,666 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:15:53,667 - INFO - machamp.model.callback - Best (16)                                                  
2025-07-10 11:15:53,667 - INFO - machamp.model.callback - dependency_las     1.5993   1.6369       0.5430     0.5572 
2025-07-10 11:15:53,667 - INFO - machamp.model.callback - lemma_accuracy     1.0484   1.0089       0.6770     0.6934 
2025-07-10 11:15:53,667 - INFO - machamp.model.callback - morph_accuracy     2.3606   2.1921       0.5432     0.5485 
2025-07-10 11:15:53,667 - INFO - machamp.model.callback - upos_accuracy      0.4214   0.3757       0.7534     0.7675 
2025-07-10 11:15:53,667 - INFO - machamp.model.callback - xpos_accuracy      0.4810   0.4329       0.7552     0.7541 
2025-07-10 11:15:53,667 - INFO - machamp.model.callback - sum                5.9107   5.6464       3.2719     3.3207 
2025-07-10 11:15:53,667 - INFO - machamp.model.callback - Epoch 16                                                   
2025-07-10 11:15:53,667 - INFO - machamp.model.callback - dependency_las     1.5993   1.6369       0.5430     0.5572 
2025-07-10 11:15:53,667 - INFO - machamp.model.callback - lemma_accuracy     1.0484   1.0089       0.6770     0.6934 
2025-07-10 11:15:53,667 - INFO - machamp.model.callback - morph_accuracy     2.3606   2.1921       0.5432     0.5485 
2025-07-10 11:15:53,667 - INFO - machamp.model.callback - upos_accuracy      0.4214   0.3757       0.7534     0.7675 
2025-07-10 11:15:53,667 - INFO - machamp.model.callback - xpos_accuracy      0.4810   0.4329       0.7552     0.7541 
2025-07-10 11:15:53,667 - INFO - machamp.model.callback - sum                5.9107   5.6464       3.2719     3.3207 
2025-07-10 11:15:53,675 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                                                   ▗▄▄▄▄▄▄▄▄│
│                                              ▄▄▞▀▀▘▀▀▘   ▗▄│
│                ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▟▀▀▀▘  ▄▄▄▄▀▀▀▀▘ │
│            ▄▄▀▀                       ▄▄▀▘▄▄▄▀▀▀▀          │ 0.6
│           ▗▘                      ▗▄▞▀▀▀▀                ▄▄│
│           ▞                  ▗▄▄▀▀▘           ▗▄▄▄▄▄▄▄▞▀▀▀▀│
│          ▗▘    ▄▄▄▄▄▄▄▗▄▄▄▀▀▀▘         ▗▄▄▄▞▄▄▄▄▞▀▀        │
│          ▞    ▐     ▗▞▘          ▗▄▄▀▀▀▘▄▞▀▀               │
│        ▄▄▌▄▄▄▄▄▄▄▄▄▞▘        ▗▄▞▀▘▗▄▄▄▞▀                   │ 0.4
│    ▄▄▀▀ ▐    ▐           ▗▄▞▀▘ ▗▄▀▘                        │
│▄▞▀▀     ▌    ▌       ▄▄▞▀▘  ▗▄▀▘                           │
│        ▐    ▐  ▄▄▄▄▄▄▄▄▄▄▄▄▛▘                              │
│        ▌▄▄▄▞▌▗▀▀▘                                          │ 0.2
│  ▗▄▄▄▗▞▀   ▐▗▘                                             │
│▀▀▘  ▗▘    ▗▞▘                                              │
│   ▄▞▘  ▄▄▀▘                                                │
│▀▀▀▀▀▀▀▀▞▘                                                  │
└────────────────────────────────────────────────────────────┘
                 5                 10                  15
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:15:53,676 - INFO - machamp.model.callback - Performance of 3.3207 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_16.pt
2025-07-10 11:15:55,485 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_15.pt
2025-07-10 11:15:55,622 - INFO - machamp.model.trainer - 
2025-07-10 11:15:55,622 - INFO - machamp.model.trainer - Epoch 17/60: training
2025-07-10 11:16:22,287 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:16:22,287 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.8140479876160993e-05
2025-07-10 11:16:22,287 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.8140479876160993e-05
2025-07-10 11:16:22,287 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 6.8933823529411776e-06
2025-07-10 11:16:22,287 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:16:22,287 - INFO - machamp.model.trainer - Epoch 17: evaluating on dev
2025-07-10 11:16:23,876 - INFO - machamp.model.callback - epoch       : 17/60
2025-07-10 11:16:23,876 - INFO - machamp.model.callback - best_epoch  : 17
2025-07-10 11:16:23,876 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:16:23,876 - INFO - machamp.model.callback - cur_ram     : 2.1276
2025-07-10 11:16:23,876 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:16:23,877 - INFO - machamp.model.callback - time_total  : 0:08:14
2025-07-10 11:16:23,877 - INFO - machamp.model.callback - 
2025-07-10 11:16:23,877 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:16:23,877 - INFO - machamp.model.callback - Best (17)                                                  
2025-07-10 11:16:23,877 - INFO - machamp.model.callback - dependency_las     1.4719   1.6280       0.5668     0.5720 
2025-07-10 11:16:23,877 - INFO - machamp.model.callback - lemma_accuracy     0.9988   0.9681       0.6925     0.7175 
2025-07-10 11:16:23,877 - INFO - machamp.model.callback - morph_accuracy     2.2164   2.0593       0.5725     0.5767 
2025-07-10 11:16:23,877 - INFO - machamp.model.callback - upos_accuracy      0.3748   0.3332       0.7867     0.8054 
2025-07-10 11:16:23,877 - INFO - machamp.model.callback - xpos_accuracy      0.4429   0.3981       0.7620     0.7612 
2025-07-10 11:16:23,877 - INFO - machamp.model.callback - sum                5.5049   5.3866       3.3805     3.4327 
2025-07-10 11:16:23,877 - INFO - machamp.model.callback - Epoch 17                                                   
2025-07-10 11:16:23,877 - INFO - machamp.model.callback - dependency_las     1.4719   1.6280       0.5668     0.5720 
2025-07-10 11:16:23,877 - INFO - machamp.model.callback - lemma_accuracy     0.9988   0.9681       0.6925     0.7175 
2025-07-10 11:16:23,877 - INFO - machamp.model.callback - morph_accuracy     2.2164   2.0593       0.5725     0.5767 
2025-07-10 11:16:23,877 - INFO - machamp.model.callback - upos_accuracy      0.3748   0.3332       0.7867     0.8054 
2025-07-10 11:16:23,877 - INFO - machamp.model.callback - xpos_accuracy      0.4429   0.3981       0.7620     0.7612 
2025-07-10 11:16:23,877 - INFO - machamp.model.callback - sum                5.5049   5.3866       3.3805     3.4327 
2025-07-10 11:16:23,885 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                                                        ▄▄▞▀│ 0.8
│                                               ▄▄▄▄▄▀▀▀▀▀▀▀▀│
│                                          ▄▄▀▀▀▀▀▀  ▗▄▄▄▞▀▀▀│
│             ▗▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▀▀▀▀▀▀▘ ▗▄▄▄▀▀▀▀▘       │
│           ▞▀▘                     ▗▄▀▘▄▀▀▀▀▘              ▗│ 0.6
│          ▗▘                   ▗▄▞▀▘▘             ▄▄▄▄▄▄▞▀▀▘│
│          ▞              ▄▄▄▞▀▀▘         ▗▄▄▟▀▄▄▄▀▀▀        │
│          ▌   ▗▀▀▀▀▀▀▄▀▀▀▀▘       ▄▄▄▞▀▀▄▄▀▀▀▀              │
│         ▐    ▞   ▄▄▀         ▄▄▀▀  ▄▄▞▀                    │ 0.4
│     ▄▄▀▀▌▀▀▀▀▀▀▀▀        ▄▄▀▀▘▗▄▀▀▀                        │
│ ▄▄▀▀   ▐    ▞        ▄▄▀▀  ▗▄▛▘                            │
│▀       ▞   ▗▘ ▄ ▗▄▞▀▀  ▗▄▄▀▘                               │
│       ▗▘▗▄▄▞▄▞▞▀▀▀▀▀▀▀▀▘                                   │ 0.2
│   ▗▄▄▗▀▀▘ ▗▗▀                                              │
│▀▀▀▘ ▄▘    ▄▘                                               │
│   ▄▞   ▄▞▀▘                                                │
│▀▀▀▀▀▀▀▀▀                                                   │
└────────────────────────────────────────────────────────────┘
                5                10                 15
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:16:23,885 - INFO - machamp.model.callback - Performance of 3.4327 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_17.pt
2025-07-10 11:16:25,660 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_16.pt
2025-07-10 11:16:25,786 - INFO - machamp.model.trainer - 
2025-07-10 11:16:25,786 - INFO - machamp.model.trainer - Epoch 18/60: training
2025-07-10 11:16:52,411 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:16:52,412 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.9235197368421053e-05
2025-07-10 11:16:52,412 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.9235197368421053e-05
2025-07-10 11:16:52,412 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 7.3093750000000015e-06
2025-07-10 11:16:52,412 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:16:52,412 - INFO - machamp.model.trainer - Epoch 18: evaluating on dev
2025-07-10 11:16:54,004 - INFO - machamp.model.callback - epoch       : 18/60
2025-07-10 11:16:54,004 - INFO - machamp.model.callback - best_epoch  : 18
2025-07-10 11:16:54,005 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:16:54,005 - INFO - machamp.model.callback - cur_ram     : 2.1360
2025-07-10 11:16:54,005 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:16:54,005 - INFO - machamp.model.callback - time_total  : 0:08:44
2025-07-10 11:16:54,005 - INFO - machamp.model.callback - 
2025-07-10 11:16:54,005 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:16:54,005 - INFO - machamp.model.callback - Best (18)                                                  
2025-07-10 11:16:54,005 - INFO - machamp.model.callback - dependency_las     1.3606   1.6580       0.5907     0.5773 
2025-07-10 11:16:54,005 - INFO - machamp.model.callback - lemma_accuracy     0.9435   0.9271       0.7206     0.7330 
2025-07-10 11:16:54,005 - INFO - machamp.model.callback - morph_accuracy     2.0762   1.9452       0.6019     0.5917 
2025-07-10 11:16:54,005 - INFO - machamp.model.callback - upos_accuracy      0.3354   0.3013       0.8159     0.8309 
2025-07-10 11:16:54,005 - INFO - machamp.model.callback - xpos_accuracy      0.4050   0.3679       0.7731     0.7789 
2025-07-10 11:16:54,005 - INFO - machamp.model.callback - sum                5.1208   5.1994       3.5022     3.5119 
2025-07-10 11:16:54,005 - INFO - machamp.model.callback - Epoch 18                                                   
2025-07-10 11:16:54,005 - INFO - machamp.model.callback - dependency_las     1.3606   1.6580       0.5907     0.5773 
2025-07-10 11:16:54,005 - INFO - machamp.model.callback - lemma_accuracy     0.9435   0.9271       0.7206     0.7330 
2025-07-10 11:16:54,005 - INFO - machamp.model.callback - morph_accuracy     2.0762   1.9452       0.6019     0.5917 
2025-07-10 11:16:54,005 - INFO - machamp.model.callback - upos_accuracy      0.3354   0.3013       0.8159     0.8309 
2025-07-10 11:16:54,005 - INFO - machamp.model.callback - xpos_accuracy      0.4050   0.3679       0.7731     0.7789 
2025-07-10 11:16:54,006 - INFO - machamp.model.callback - sum                5.1208   5.1994       3.5022     3.5119 
2025-07-10 11:16:54,014 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                                                       ▄▄▄▀▀│
│                                                ▗▄▄▄▄▄▄▄▄▄▀▀│ 0.8
│                                         ▗▄▄▀▀▀▀▘    ▄▄▄▞▀▀▀│
│              ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▀▀▘▀ ▗▄▄▄▞▀▀▀▀       │
│          ▗▄▞▀▘                   ▗▄▀▘▄▄▄▀▀▀▘               │
│          ▞                   ▗▄▞▀▘▀▀             ▗▄▄▄▄▞▀▀▀▀│
│          ▌               ▄▄▄▀▀          ▄▄▄▄▄▄▄▀▀▀▀        │ 0.5
│         ▐    ▛▀▀▀▀▀▗▄▀▀▀▀▘       ▄▄▄▞▗▄▄▄▞▀▀               │
│         ▌   ▐    ▄▞▘        ▄▄▀▀▀ ▗▄▀▘                     │
│     ▗▄▄▐▀▀▀▀▀▀▀▀▀       ▄▄▀▀ ▗▞▀▀▀▘                        │
│ ▗▄▞▀▘  ▞   ▐        ▗▄▞▀  ▗▄▀▘                             │
│▀▘     ▗▘   ▌    ▄▄▀▀▘  ▗▄▀▘                                │ 0.2
│       ▞  ▄▐▄▄▀▀▀▀▀▀▀▀▀▀▘                                   │
│    ▄▄▄▘▀▀ ▞▞                                               │
│▀▀▀▀ ▞    ▗▞                                                │
│   ▄▞  ▗▄▀▀                                                 │
│▀▀▀▀▀▀▀▘▘                                                   │
└────────────────────────────────────────────────────────────┘
               5               10                15
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:16:54,014 - INFO - machamp.model.callback - Performance of 3.5119 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_18.pt
2025-07-10 11:16:55,793 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_17.pt
2025-07-10 11:16:55,922 - INFO - machamp.model.trainer - 
2025-07-10 11:16:55,923 - INFO - machamp.model.trainer - Epoch 19/60: training
2025-07-10 11:17:22,537 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:17:22,537 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.985862354892206e-05
2025-07-10 11:17:22,537 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.985862354892206e-05
2025-07-10 11:17:22,537 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 7.546276948590383e-06
2025-07-10 11:17:22,538 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:17:22,538 - INFO - machamp.model.trainer - Epoch 19: evaluating on dev
2025-07-10 11:17:24,130 - INFO - machamp.model.callback - epoch       : 19/60
2025-07-10 11:17:24,130 - INFO - machamp.model.callback - best_epoch  : 19
2025-07-10 11:17:24,130 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:17:24,130 - INFO - machamp.model.callback - cur_ram     : 2.1406
2025-07-10 11:17:24,130 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:17:24,130 - INFO - machamp.model.callback - time_total  : 0:09:14
2025-07-10 11:17:24,130 - INFO - machamp.model.callback - 
2025-07-10 11:17:24,130 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:17:24,130 - INFO - machamp.model.callback - Best (19)                                                  
2025-07-10 11:17:24,130 - INFO - machamp.model.callback - dependency_las     1.2664   1.6223       0.6063     0.5793 
2025-07-10 11:17:24,130 - INFO - machamp.model.callback - lemma_accuracy     0.8961   0.8881       0.7354     0.7491 
2025-07-10 11:17:24,130 - INFO - machamp.model.callback - morph_accuracy     1.9423   1.8276       0.6257     0.6095 
2025-07-10 11:17:24,130 - INFO - machamp.model.callback - upos_accuracy      0.2988   0.2705       0.8357     0.8497 
2025-07-10 11:17:24,131 - INFO - machamp.model.callback - xpos_accuracy      0.3721   0.3350       0.7899     0.7960 
2025-07-10 11:17:24,131 - INFO - machamp.model.callback - sum                4.7757   4.9435       3.5930     3.5837 
2025-07-10 11:17:24,131 - INFO - machamp.model.callback - Epoch 19                                                   
2025-07-10 11:17:24,131 - INFO - machamp.model.callback - dependency_las     1.2664   1.6223       0.6063     0.5793 
2025-07-10 11:17:24,131 - INFO - machamp.model.callback - lemma_accuracy     0.8961   0.8881       0.7354     0.7491 
2025-07-10 11:17:24,131 - INFO - machamp.model.callback - morph_accuracy     1.9423   1.8276       0.6257     0.6095 
2025-07-10 11:17:24,131 - INFO - machamp.model.callback - upos_accuracy      0.2988   0.2705       0.8357     0.8497 
2025-07-10 11:17:24,131 - INFO - machamp.model.callback - xpos_accuracy      0.3721   0.3350       0.7899     0.7960 
2025-07-10 11:17:24,131 - INFO - machamp.model.callback - sum                4.7757   4.9435       3.5930     3.5837 
2025-07-10 11:17:24,139 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                                                      ▄▄▄▞▀▀│
│                                                 ▄▄▞▀▄▄▄▄▄▀▀│ 0.8
│                                        ▗▄▄▞▀▀▀▀▀▀▀▀▀▄▄▄▄▀▀▀│
│                                   ▄▄▄▞▀▘▀▀▘▗▄▄▄▞▀▀▀        │
│           ▄▞▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▘▄▄▄▞▀▀▀▘              ▗│
│         ▗▀                  ▗▄▄▀▘▀▀▘            ▗▄▄▄▄▞▀▀▀▀▘│ 0.6
│         ▐                ▄▄▀▀▘         ▗▄▄▄▄▄▄▀▀▀▀         │
│         ▌   ▄▄▄▄▄▄▗▄▄▞▀▀▀       ▗▄▄▄▄▄▄▄▀▀▀                │
│        ▐    ▌   ▗▞▘        ▄▄▛▀▀▘▄▞▀                       │
│     ▗▄▄▞▄▄▄▄▄▞▀▀▘      ▗▄▞▀ ▄▄▀▀▀                          │ 0.4
│  ▄▄▀▘ ▗▘   ▌       ▗▄▞▀▘ ▗▞▀                               │
│▀▀     ▐   ▐    ▗▄▞▀▘  ▗▄▀▘                                 │
│       ▌  ▄▌▗▀▀▀▀▀▀▀▀▀▀▘                                    │
│    ▄▄▞▀▀▀▐▗▘                                               │ 0.2
│▀▀▀▀▗▞    ▗▘                                                │
│   ▄▘  ▄▞▀▘                                                 │
│▀▀▀▀▀▀▀▀                                                    │
└────────────────────────────────────────────────────────────┘
              5               10              15
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:17:24,139 - INFO - machamp.model.callback - Performance of 3.5837 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_19.pt
2025-07-10 11:17:25,881 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_18.pt
2025-07-10 11:17:26,011 - INFO - machamp.model.trainer - 
2025-07-10 11:17:26,011 - INFO - machamp.model.trainer - Epoch 20/60: training
2025-07-10 11:17:52,599 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:17:52,599 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.9389510779436156e-05
2025-07-10 11:17:52,599 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.9389510779436156e-05
2025-07-10 11:17:52,599 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 7.36801409618574e-06
2025-07-10 11:17:52,599 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:17:52,599 - INFO - machamp.model.trainer - Epoch 20: evaluating on dev
2025-07-10 11:17:54,178 - INFO - machamp.model.callback - epoch       : 20/60
2025-07-10 11:17:54,179 - INFO - machamp.model.callback - best_epoch  : 20
2025-07-10 11:17:54,179 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:17:54,179 - INFO - machamp.model.callback - cur_ram     : 2.1406
2025-07-10 11:17:54,179 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:17:54,179 - INFO - machamp.model.callback - time_total  : 0:09:44
2025-07-10 11:17:54,179 - INFO - machamp.model.callback - 
2025-07-10 11:17:54,179 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:17:54,179 - INFO - machamp.model.callback - Best (20)                                                  
2025-07-10 11:17:54,179 - INFO - machamp.model.callback - dependency_las     1.1447   1.6690       0.6348     0.5941 
2025-07-10 11:17:54,179 - INFO - machamp.model.callback - lemma_accuracy     0.8555   0.8618       0.7472     0.7601 
2025-07-10 11:17:54,179 - INFO - machamp.model.callback - morph_accuracy     1.8384   1.7374       0.6409     0.6189 
2025-07-10 11:17:54,179 - INFO - machamp.model.callback - upos_accuracy      0.2729   0.2476       0.8554     0.8578 
2025-07-10 11:17:54,179 - INFO - machamp.model.callback - xpos_accuracy      0.3432   0.3139       0.8050     0.8085 
2025-07-10 11:17:54,179 - INFO - machamp.model.callback - sum                4.4547   4.8297       3.6834     3.6394 
2025-07-10 11:17:54,179 - INFO - machamp.model.callback - Epoch 20                                                   
2025-07-10 11:17:54,179 - INFO - machamp.model.callback - dependency_las     1.1447   1.6690       0.6348     0.5941 
2025-07-10 11:17:54,179 - INFO - machamp.model.callback - lemma_accuracy     0.8555   0.8618       0.7472     0.7601 
2025-07-10 11:17:54,180 - INFO - machamp.model.callback - morph_accuracy     1.8384   1.7374       0.6409     0.6189 
2025-07-10 11:17:54,180 - INFO - machamp.model.callback - upos_accuracy      0.2729   0.2476       0.8554     0.8578 
2025-07-10 11:17:54,180 - INFO - machamp.model.callback - xpos_accuracy      0.3432   0.3139       0.8050     0.8085 
2025-07-10 11:17:54,180 - INFO - machamp.model.callback - sum                4.4547   4.8297       3.6834     3.6394 
2025-07-10 11:17:54,188 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                                                    ▄▄▄▄▀▀▀▀│
│                                               ▗▄▞▀▗▄▄▄▄▀▀▀▀│ 0.8
│                                       ▄▄▄▄▀▀▀▀▀▀▀▀▘▄▄▄▞▀▀▀▀│
│                                  ▗▄▄▀▀▀▀▘ ▄▄▄▄▀▀▀          │
│           ▄▞▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▗▄▄▄▞▀▀▀               ▄▄│
│         ▐▀                 ▗▄▞▀▀▀▀▘            ▗▄▄▄▞▀▀▀▀▀▞▀│ 0.6
│         ▌               ▄▄▀▀          ▄▄▄▄▄▄▞▀▀▘           │
│        ▐   ▗▄▄▄▄▄▄▄▄▄▀▀▀       ▗▄▄▞▄▄▄▞▀▀                  │
│        ▞   ▞    ▄▀        ▄▄▀▀▀▗▞▀▀                        │
│     ▗▄▗▘▄▄▄▄▄▄▀▀      ▗▄▀▀▗▄▄▀▀▘                           │ 0.4
│  ▄▄▀▘ ▐   ▞        ▄▞▀▘ ▄▞▘                                │
│▀▀     ▌   ▌   ▗▄▄▀▀  ▗▄▀                                   │
│      ▐  ▗▐▄▞▀▀▀▀▀▀▀▀▀▘                                     │
│    ▄▗▞▀▀▘▌▞                                                │ 0.2
│▞▀▀▀▗▘   ▐▞                                                 │
│   ▄▘ ▗▄▞▀                                                  │
│▀▀▀▀▀▀▀▘                                                    │
└────────────────────────────────────────────────────────────┘
             5              10              15             20
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:17:54,188 - INFO - machamp.model.callback - Performance of 3.6394 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_20.pt
2025-07-10 11:17:55,965 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_19.pt
2025-07-10 11:17:56,103 - INFO - machamp.model.trainer - 
2025-07-10 11:17:56,103 - INFO - machamp.model.trainer - Epoch 21/60: training
2025-07-10 11:18:22,603 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:18:22,603 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.892039800995025e-05
2025-07-10 11:18:22,603 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.892039800995025e-05
2025-07-10 11:18:22,603 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 7.189751243781095e-06
2025-07-10 11:18:22,603 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:18:22,604 - INFO - machamp.model.trainer - Epoch 21: evaluating on dev
2025-07-10 11:18:24,168 - INFO - machamp.model.callback - epoch       : 21/60
2025-07-10 11:18:24,168 - INFO - machamp.model.callback - best_epoch  : 21
2025-07-10 11:18:24,169 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:18:24,169 - INFO - machamp.model.callback - cur_ram     : 2.1406
2025-07-10 11:18:24,169 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:18:24,169 - INFO - machamp.model.callback - time_total  : 0:10:14
2025-07-10 11:18:24,169 - INFO - machamp.model.callback - 
2025-07-10 11:18:24,169 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:18:24,169 - INFO - machamp.model.callback - Best (21)                                                  
2025-07-10 11:18:24,169 - INFO - machamp.model.callback - dependency_las     1.0642   1.6658       0.6521     0.6035 
2025-07-10 11:18:24,169 - INFO - machamp.model.callback - lemma_accuracy     0.8199   0.8385       0.7649     0.7695 
2025-07-10 11:18:24,169 - INFO - machamp.model.callback - morph_accuracy     1.7276   1.6539       0.6573     0.6327 
2025-07-10 11:18:24,169 - INFO - machamp.model.callback - upos_accuracy      0.2436   0.2273       0.8729     0.8698 
2025-07-10 11:18:24,169 - INFO - machamp.model.callback - xpos_accuracy      0.3164   0.2910       0.8243     0.8232 
2025-07-10 11:18:24,169 - INFO - machamp.model.callback - sum                4.1718   4.6765       3.7715     3.6988 
2025-07-10 11:18:24,169 - INFO - machamp.model.callback - Epoch 21                                                   
2025-07-10 11:18:24,170 - INFO - machamp.model.callback - dependency_las     1.0642   1.6658       0.6521     0.6035 
2025-07-10 11:18:24,170 - INFO - machamp.model.callback - lemma_accuracy     0.8199   0.8385       0.7649     0.7695 
2025-07-10 11:18:24,170 - INFO - machamp.model.callback - morph_accuracy     1.7276   1.6539       0.6573     0.6327 
2025-07-10 11:18:24,170 - INFO - machamp.model.callback - upos_accuracy      0.2436   0.2273       0.8729     0.8698 
2025-07-10 11:18:24,170 - INFO - machamp.model.callback - xpos_accuracy      0.3164   0.2910       0.8243     0.8232 
2025-07-10 11:18:24,170 - INFO - machamp.model.callback - sum                4.1718   4.6765       3.7715     3.6988 
2025-07-10 11:18:24,178 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                                                   ▄▄▄▄▄▞▀▀▀│
│                                              ▄▟▀▀▀▄▄▄▄▄▀▀▀▀│ 0.8
│                                      ▄▄▄▄▄▄▞▀▀▀▀▀▀▄▄▄▄▀▀▀▀▀│
│                                 ▗▄▞▀▀▀▀▘ ▄▄▄▞▀▀▀▘          │
│           ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▀▀▀▀▘ ▄▄▄▞▀▀▀               ▗▄▄│
│         ▛▀                 ▄▞▀▞▀▀▀            ▄▄▄▄▞▀▀▀▀▀▘▀▀│ 0.6
│        ▐               ▗▄▀▀▘         ▗▄▗▄▄▄▀▀▀▘            │
│        ▞   ▄▄▄▄▄▄▗▄▄▞▀▀▘      ▗▄▄▞▗▄▄▀▀▘                   │
│       ▗▘  ▗▘   ▄▀▘       ▄▄▞▀▀▄▞▀▀▘                        │
│      ▄▐▄▄▄▄▄▄▄▛      ▗▄▞▀▗▄▄▞▀                             │ 0.4
│  ▄▄▀▀ ▌  ▗▘       ▄▄▀▘ ▄▞▘                                 │
│▞▀    ▗▘  ▐    ▗▄▞▀  ▗▄▀▘                                   │
│      ▞  ▗▌▗▀▀▀▀▀▚▛▀▀▘                                      │
│    ▄▗▘▀▀▐▗▘                                                │ 0.2
│▞▀▀▀▗▘   ▗▘                                                 │
│   ▄▘ ▗▄▀▘                                                  │
│▀▀▀▀▀▀▘▘                                                    │
└────────────────────────────────────────────────────────────┘
             5             10            15             20
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:18:24,178 - INFO - machamp.model.callback - Performance of 3.6988 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_21.pt
2025-07-10 11:18:25,936 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_20.pt
2025-07-10 11:18:26,072 - INFO - machamp.model.trainer - 
2025-07-10 11:18:26,072 - INFO - machamp.model.trainer - Epoch 22/60: training
2025-07-10 11:18:52,557 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:18:52,557 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.8451285240464346e-05
2025-07-10 11:18:52,557 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.8451285240464346e-05
2025-07-10 11:18:52,557 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 7.0114883913764516e-06
2025-07-10 11:18:52,557 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:18:52,557 - INFO - machamp.model.trainer - Epoch 22: evaluating on dev
2025-07-10 11:18:54,173 - INFO - machamp.model.callback - epoch       : 22/60
2025-07-10 11:18:54,173 - INFO - machamp.model.callback - best_epoch  : 22
2025-07-10 11:18:54,173 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:18:54,173 - INFO - machamp.model.callback - cur_ram     : 2.1407
2025-07-10 11:18:54,173 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:18:54,173 - INFO - machamp.model.callback - time_total  : 0:10:44
2025-07-10 11:18:54,173 - INFO - machamp.model.callback - 
2025-07-10 11:18:54,173 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:18:54,173 - INFO - machamp.model.callback - Best (22)                                                  
2025-07-10 11:18:54,174 - INFO - machamp.model.callback - dependency_las     0.9753   1.6955       0.6673     0.6082 
2025-07-10 11:18:54,174 - INFO - machamp.model.callback - lemma_accuracy     0.7831   0.8094       0.7732     0.7789 
2025-07-10 11:18:54,174 - INFO - machamp.model.callback - morph_accuracy     1.6365   1.5766       0.6682     0.6441 
2025-07-10 11:18:54,174 - INFO - machamp.model.callback - upos_accuracy      0.2212   0.2140       0.8803     0.8772 
2025-07-10 11:18:54,174 - INFO - machamp.model.callback - xpos_accuracy      0.2939   0.2747       0.8388     0.8309 
2025-07-10 11:18:54,174 - INFO - machamp.model.callback - sum                3.9100   4.5703       3.8278     3.7393 
2025-07-10 11:18:54,174 - INFO - machamp.model.callback - Epoch 22                                                   
2025-07-10 11:18:54,174 - INFO - machamp.model.callback - dependency_las     0.9753   1.6955       0.6673     0.6082 
2025-07-10 11:18:54,174 - INFO - machamp.model.callback - lemma_accuracy     0.7831   0.8094       0.7732     0.7789 
2025-07-10 11:18:54,174 - INFO - machamp.model.callback - morph_accuracy     1.6365   1.5766       0.6682     0.6441 
2025-07-10 11:18:54,174 - INFO - machamp.model.callback - upos_accuracy      0.2212   0.2140       0.8803     0.8772 
2025-07-10 11:18:54,174 - INFO - machamp.model.callback - xpos_accuracy      0.2939   0.2747       0.8388     0.8309 
2025-07-10 11:18:54,174 - INFO - machamp.model.callback - sum                3.9100   4.5703       3.8278     3.7393 
2025-07-10 11:18:54,183 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                                                 ▗▄▄▄▄▄▞▀▀▀▀│
│                                            ▄▄▞▀▀▗▄▄▄▄▞▀▀▀▀▀│ 0.8
│                                     ▄▄▄▄▄▄▄▄▀▀▀▀▘▄▄▄▄▀▀▀▀▀▀│
│                                ▗▄▞▀▀▀▀  ▄▄▄▞▀▀▀▘           │
│          ▗▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▀▀▀▘▘▗▄▄▞▀▀▀               ▄▄▄▄│
│        ▗▀▘                ▄▞▘▞▀▀▀▘          ▗▄▄▄▄▀▀▀▀▀▀▀▀▀▀│ 0.6
│        ▞              ▗▄▞▀▘         ▗▄▄▄▄▞▀▀▘▀             │
│       ▗▘  ▗▄▄▄▄▄▄▄▄▄▀▀▘      ▗▄▄▞▗▄▞▀▀                     │
│       ▐   ▞   ▗▞▀       ▄▄▞▀▀▄▀▀▀▘                         │
│     ▗▄▌▄▄▄▄▄▄▞▘     ▗▄▞▀▗▄▄▞▀                              │ 0.4
│  ▗▄▀▘▗▘  ▞       ▗▄▀▘ ▄▞▘                                  │
│▞▀▘   ▐   ▌   ▗▄▞▀▘ ▗▄▀                                     │
│      ▌  ▐▄▞▀▀▙▄▄▞▀▀▘                                       │
│    ▄▟▞▀▀▌▞                                                 │ 0.2
│▞▀▀▀▞   ▗▞                                                  │
│  ▗▞  ▄▞▀                                                   │
│▀▀▘▀▀▀▀                                                     │
└────────────────────────────────────────────────────────────┘
            5            10            15             20
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:18:54,183 - INFO - machamp.model.callback - Performance of 3.7393 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_22.pt
2025-07-10 11:18:55,956 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_21.pt
2025-07-10 11:18:56,085 - INFO - machamp.model.trainer - 
2025-07-10 11:18:56,085 - INFO - machamp.model.trainer - Epoch 23/60: training
2025-07-10 11:19:22,523 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:19:22,523 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.7982172470978442e-05
2025-07-10 11:19:22,524 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.7982172470978442e-05
2025-07-10 11:19:22,524 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 6.833225538971809e-06
2025-07-10 11:19:22,524 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:19:22,524 - INFO - machamp.model.trainer - Epoch 23: evaluating on dev
2025-07-10 11:19:24,099 - INFO - machamp.model.callback - epoch       : 23/60
2025-07-10 11:19:24,100 - INFO - machamp.model.callback - best_epoch  : 23
2025-07-10 11:19:24,100 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:19:24,100 - INFO - machamp.model.callback - cur_ram     : 2.1277
2025-07-10 11:19:24,100 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:19:24,100 - INFO - machamp.model.callback - time_total  : 0:11:14
2025-07-10 11:19:24,100 - INFO - machamp.model.callback - 
2025-07-10 11:19:24,100 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:19:24,100 - INFO - machamp.model.callback - Best (23)                                                  
2025-07-10 11:19:24,100 - INFO - machamp.model.callback - dependency_las     0.8796   1.7179       0.6923     0.6169 
2025-07-10 11:19:24,100 - INFO - machamp.model.callback - lemma_accuracy     0.7546   0.7914       0.7791     0.7840 
2025-07-10 11:19:24,100 - INFO - machamp.model.callback - morph_accuracy     1.5585   1.5185       0.6837     0.6511 
2025-07-10 11:19:24,100 - INFO - machamp.model.callback - upos_accuracy      0.2050   0.2026       0.8923     0.8782 
2025-07-10 11:19:24,100 - INFO - machamp.model.callback - xpos_accuracy      0.2740   0.2590       0.8495     0.8397 
2025-07-10 11:19:24,100 - INFO - machamp.model.callback - sum                3.6718   4.4893       3.8970     3.7699 
2025-07-10 11:19:24,100 - INFO - machamp.model.callback - Epoch 23                                                   
2025-07-10 11:19:24,100 - INFO - machamp.model.callback - dependency_las     0.8796   1.7179       0.6923     0.6169 
2025-07-10 11:19:24,101 - INFO - machamp.model.callback - lemma_accuracy     0.7546   0.7914       0.7791     0.7840 
2025-07-10 11:19:24,101 - INFO - machamp.model.callback - morph_accuracy     1.5585   1.5185       0.6837     0.6511 
2025-07-10 11:19:24,101 - INFO - machamp.model.callback - upos_accuracy      0.2050   0.2026       0.8923     0.8782 
2025-07-10 11:19:24,101 - INFO - machamp.model.callback - xpos_accuracy      0.2740   0.2590       0.8495     0.8397 
2025-07-10 11:19:24,101 - INFO - machamp.model.callback - sum                3.6718   4.4893       3.8970     3.7699 
2025-07-10 11:19:24,110 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                                               ▗▄▄▄▄▄▞▀▀▀▀▀▗│
│                                          ▄▄▞▀▀▗▄▄▄▄▞▀▀▀▀▀▀▘│ 0.8
│                                   ▄▄▄▄▄▄▄▄▀▀▀▀▘▄▄▄▟▀▀▀▀▀▀▀▀│
│                               ▄▄▀▀▘▀  ▄▄▄▞▀▀▀▘             │
│          ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▀▀▀▀ ▄▄▞▀▀▀              ▗▄▄▄▄▄▄│
│        ▞▀               ▗▄▀▘▞▀▀▀          ▄▄▄▄▞▀▀▀▀▀▘▀▀▀▀▀▀│ 0.6
│        ▌             ▗▄▀▘▘         ▄▗▄▄▄▀▀▘▀               │
│       ▐   ▄▄▄▄▄▄▄▄▄▀▀▘      ▄▄▄▀▄▄▀▀▘                      │
│       ▌  ▐    ▄▀       ▄▄▀▀▗▞▀▀▀                           │
│     ▗▗▘▄▄▄▄▄▄▀      ▄▞▀▗▄▄▀▘                               │ 0.4
│  ▄▄▀▘▐   ▌      ▗▄▀▀ ▄▀▘                                   │
│▞▀    ▌  ▐    ▄▄▀▘ ▗▄▀                                      │
│     ▗▘  ▌▗▀▀▚▄▄▄▀▀▘                                        │
│    ▄▞▀▀▗▗▘                                                 │ 0.2
│▞▀▀▀▞   ▗▘                                                  │
│  ▗▞  ▄▞▘                                                   │
│▀▀▘▀▀▀▘                                                     │
└────────────────────────────────────────────────────────────┘
           5            10            15           20
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:19:24,110 - INFO - machamp.model.callback - Performance of 3.7699 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_23.pt
2025-07-10 11:19:25,883 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_22.pt
2025-07-10 11:19:26,010 - INFO - machamp.model.trainer - 
2025-07-10 11:19:26,010 - INFO - machamp.model.trainer - Epoch 24/60: training
2025-07-10 11:19:52,543 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:19:52,544 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.751305970149254e-05
2025-07-10 11:19:52,544 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.751305970149254e-05
2025-07-10 11:19:52,544 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 6.654962686567165e-06
2025-07-10 11:19:52,544 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:19:52,544 - INFO - machamp.model.trainer - Epoch 24: evaluating on dev
2025-07-10 11:19:54,197 - INFO - machamp.model.callback - epoch       : 24/60
2025-07-10 11:19:54,198 - INFO - machamp.model.callback - best_epoch  : 24
2025-07-10 11:19:54,198 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:19:54,198 - INFO - machamp.model.callback - cur_ram     : 2.1277
2025-07-10 11:19:54,198 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:19:54,198 - INFO - machamp.model.callback - time_total  : 0:11:44
2025-07-10 11:19:54,198 - INFO - machamp.model.callback - 
2025-07-10 11:19:54,198 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:19:54,198 - INFO - machamp.model.callback - Best (24)                                                  
2025-07-10 11:19:54,198 - INFO - machamp.model.callback - dependency_las     0.8201   1.7003       0.7058     0.6129 
2025-07-10 11:19:54,198 - INFO - machamp.model.callback - lemma_accuracy     0.7285   0.7693       0.7854     0.7944 
2025-07-10 11:19:54,198 - INFO - machamp.model.callback - morph_accuracy     1.4873   1.4619       0.6930     0.6598 
2025-07-10 11:19:54,198 - INFO - machamp.model.callback - upos_accuracy      0.1876   0.1903       0.8989     0.8849 
2025-07-10 11:19:54,199 - INFO - machamp.model.callback - xpos_accuracy      0.2567   0.2456       0.8593     0.8457 
2025-07-10 11:19:54,199 - INFO - machamp.model.callback - sum                3.4802   4.3674       3.9424     3.7977 
2025-07-10 11:19:54,199 - INFO - machamp.model.callback - Epoch 24                                                   
2025-07-10 11:19:54,199 - INFO - machamp.model.callback - dependency_las     0.8201   1.7003       0.7058     0.6129 
2025-07-10 11:19:54,199 - INFO - machamp.model.callback - lemma_accuracy     0.7285   0.7693       0.7854     0.7944 
2025-07-10 11:19:54,199 - INFO - machamp.model.callback - morph_accuracy     1.4873   1.4619       0.6930     0.6598 
2025-07-10 11:19:54,199 - INFO - machamp.model.callback - upos_accuracy      0.1876   0.1903       0.8989     0.8849 
2025-07-10 11:19:54,199 - INFO - machamp.model.callback - xpos_accuracy      0.2567   0.2456       0.8593     0.8457 
2025-07-10 11:19:54,199 - INFO - machamp.model.callback - sum                3.4802   4.3674       3.9424     3.7977 
2025-07-10 11:19:54,207 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                                              ▗▄▄▄▄▄▄▀▀▀▀▀▀▀│
│                                        ▗▄▄▀▀▀▗▄▄▄▄▞▀▀▀▀▀▀▀▀│ 0.8
│                                   ▄▄▄▄▄▄▄▞▀▀▀▘▄▄▄▄▀▀▀▀▀▀▀▀▘│
│                              ▄▄▀▀▀▀  ▄▄▄▀▀▀▀               │
│          ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▞▀▀▀ ▄▄▄▀▀▀              ▄▄▄▄▄▄▄▞│
│        ▞▀               ▄▛▘▞▀▀▀          ▄▄▄▄▞▀▀▀▀▀▄▞▀▀▀▀▀▀│ 0.6
│       ▐             ▗▄▞▀▀         ▄▄▄▄▞▀▀▀▀▀▀▀             │
│       ▞         ▄▄▞▀▀      ▗▄▄▗▄▟▀▀▘                       │
│       ▌  ▞▀▀▀▗▞▀      ▄▄▞▀▗▞▀▀▘                            │
│     ▄▐▄▄▄▄▄▄▞▘     ▄▞▀▗▄▄▛▘                                │ 0.4
│  ▗▄▀▘▞  ▐       ▄▞▀ ▄▞▘                                    │
│▞▀▘   ▌  ▌   ▗▄▞▀  ▄▀                                       │
│     ▐  ▗▘▞▀▚▄▄▄▞▀▀                                         │
│    ▄▛▀▀▞▞                                                  │ 0.2
│▄▀▀▀▞   ▞                                                   │
│  ▗▞ ▗▄▀▘                                                   │
│▀▀▘▀▀▘▘                                                     │
└────────────────────────────────────────────────────────────┘
           5           10           15           20
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:19:54,208 - INFO - machamp.model.callback - Performance of 3.7977 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_24.pt
2025-07-10 11:19:55,982 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_23.pt
2025-07-10 11:19:56,108 - INFO - machamp.model.trainer - 
2025-07-10 11:19:56,108 - INFO - machamp.model.trainer - Epoch 25/60: training
2025-07-10 11:20:22,567 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:20:22,567 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.7043946932006636e-05
2025-07-10 11:20:22,567 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.7043946932006636e-05
2025-07-10 11:20:22,567 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 6.476699834162522e-06
2025-07-10 11:20:22,567 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:20:22,567 - INFO - machamp.model.trainer - Epoch 25: evaluating on dev
2025-07-10 11:20:24,130 - INFO - machamp.model.callback - epoch       : 25/60
2025-07-10 11:20:24,130 - INFO - machamp.model.callback - best_epoch  : 25
2025-07-10 11:20:24,130 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:20:24,131 - INFO - machamp.model.callback - cur_ram     : 2.1277
2025-07-10 11:20:24,131 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:20:24,131 - INFO - machamp.model.callback - time_total  : 0:12:14
2025-07-10 11:20:24,131 - INFO - machamp.model.callback - 
2025-07-10 11:20:24,131 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:20:24,131 - INFO - machamp.model.callback - Best (25)                                                  
2025-07-10 11:20:24,131 - INFO - machamp.model.callback - dependency_las     0.7375   1.7562       0.7265     0.6132 
2025-07-10 11:20:24,131 - INFO - machamp.model.callback - lemma_accuracy     0.7030   0.7533       0.7916     0.8021 
2025-07-10 11:20:24,131 - INFO - machamp.model.callback - morph_accuracy     1.4241   1.4100       0.7014     0.6743 
2025-07-10 11:20:24,131 - INFO - machamp.model.callback - upos_accuracy      0.1754   0.1835       0.9069     0.8896 
2025-07-10 11:20:24,131 - INFO - machamp.model.callback - xpos_accuracy      0.2422   0.2365       0.8675     0.8507 
2025-07-10 11:20:24,131 - INFO - machamp.model.callback - sum                3.2822   4.3395       3.9939     3.8299 
2025-07-10 11:20:24,131 - INFO - machamp.model.callback - Epoch 25                                                   
2025-07-10 11:20:24,131 - INFO - machamp.model.callback - dependency_las     0.7375   1.7562       0.7265     0.6132 
2025-07-10 11:20:24,131 - INFO - machamp.model.callback - lemma_accuracy     0.7030   0.7533       0.7916     0.8021 
2025-07-10 11:20:24,131 - INFO - machamp.model.callback - morph_accuracy     1.4241   1.4100       0.7014     0.6743 
2025-07-10 11:20:24,131 - INFO - machamp.model.callback - upos_accuracy      0.1754   0.1835       0.9069     0.8896 
2025-07-10 11:20:24,131 - INFO - machamp.model.callback - xpos_accuracy      0.2422   0.2365       0.8675     0.8507 
2025-07-10 11:20:24,131 - INFO - machamp.model.callback - sum                3.2822   4.3395       3.9939     3.8299 
2025-07-10 11:20:24,140 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                                             ▄▄▄▄▄▄▄▀▀▀▀▀▀▀▗│
│                                       ▗▄▞▀▀▀▄▄▄▄▄▀▀▀▀▀▀▀▀▀▘│ 0.8
│                                  ▄▄▄▄▄▄▄▀▀▀▀▄▄▄▄▄▀▀▀▀▀▀▀▀  │
│                             ▄▄▀▀▀▀  ▄▄▟▀▀▀▀                │
│          ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▀▀▀▘▄▄▄▀▀▀             ▗▄▄▄▄▄▄▄▀▀│
│       ▗▞▀              ▄▞▘▟▀▀▀          ▄▄▄▟▀▀▀▀▀▘▄▞▀▀▀▀▀▀▀│ 0.6
│       ▞             ▄▟▀▀▘       ▗▄▄▄▄▀▀▀▀▀▀▀▘              │
│       ▌        ▗▄▄▀▀      ▗▄▄▗▄▞▀▀                         │
│      ▐  ▗▀▀▀▀▞▀▘▀    ▗▄▞▀▗▞▀▀▘                             │
│     ▗▞▄▄▄▄▄▄▀     ▗▄▀▗▄▄▛▘                                 │ 0.4
│  ▗▄▀▘▌  ▌      ▗▄▀▘▗▞▘                                     │
│▄▀▘  ▐  ▗▘   ▄▄▀▘ ▄▞▘                                       │
│     ▞  ▐▗▀▚▄▄▄▄▀▀                                          │
│   ▗▗▘▀▀▗▘                                                  │ 0.2
│▄▀▀▗▘  ▐▞                                                   │
│  ▗▘ ▗▄▀                                                    │
│▀▀▘▀▀▘                                                      │
└────────────────────────────────────────────────────────────┘
           5          10          15           20          25
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:20:24,140 - INFO - machamp.model.callback - Performance of 3.8299 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_25.pt
2025-07-10 11:20:25,991 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_24.pt
2025-07-10 11:20:26,118 - INFO - machamp.model.trainer - 
2025-07-10 11:20:26,119 - INFO - machamp.model.trainer - Epoch 26/60: training
2025-07-10 11:20:52,732 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:20:52,732 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.657483416252073e-05
2025-07-10 11:20:52,732 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.657483416252073e-05
2025-07-10 11:20:52,732 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 6.2984369817578776e-06
2025-07-10 11:20:52,732 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:20:52,732 - INFO - machamp.model.trainer - Epoch 26: evaluating on dev
2025-07-10 11:20:54,321 - INFO - machamp.model.callback - epoch       : 26/60
2025-07-10 11:20:54,321 - INFO - machamp.model.callback - best_epoch  : 26
2025-07-10 11:20:54,322 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:20:54,322 - INFO - machamp.model.callback - cur_ram     : 2.1278
2025-07-10 11:20:54,322 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:20:54,322 - INFO - machamp.model.callback - time_total  : 0:12:44
2025-07-10 11:20:54,322 - INFO - machamp.model.callback - 
2025-07-10 11:20:54,322 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:20:54,322 - INFO - machamp.model.callback - Best (26)                                                  
2025-07-10 11:20:54,322 - INFO - machamp.model.callback - dependency_las     0.7215   1.7908       0.7339     0.6290 
2025-07-10 11:20:54,322 - INFO - machamp.model.callback - lemma_accuracy     0.6780   0.7380       0.8005     0.8068 
2025-07-10 11:20:54,322 - INFO - machamp.model.callback - morph_accuracy     1.3574   1.3667       0.7123     0.6806 
2025-07-10 11:20:54,322 - INFO - machamp.model.callback - upos_accuracy      0.1607   0.1763       0.9121     0.8916 
2025-07-10 11:20:54,322 - INFO - machamp.model.callback - xpos_accuracy      0.2274   0.2283       0.8747     0.8574 
2025-07-10 11:20:54,322 - INFO - machamp.model.callback - sum                3.1450   4.3001       4.0334     3.8655 
2025-07-10 11:20:54,322 - INFO - machamp.model.callback - Epoch 26                                                   
2025-07-10 11:20:54,322 - INFO - machamp.model.callback - dependency_las     0.7215   1.7908       0.7339     0.6290 
2025-07-10 11:20:54,322 - INFO - machamp.model.callback - lemma_accuracy     0.6780   0.7380       0.8005     0.8068 
2025-07-10 11:20:54,322 - INFO - machamp.model.callback - morph_accuracy     1.3574   1.3667       0.7123     0.6806 
2025-07-10 11:20:54,322 - INFO - machamp.model.callback - upos_accuracy      0.1607   0.1763       0.9121     0.8916 
2025-07-10 11:20:54,322 - INFO - machamp.model.callback - xpos_accuracy      0.2274   0.2283       0.8747     0.8574 
2025-07-10 11:20:54,322 - INFO - machamp.model.callback - sum                3.1450   4.3001       4.0334     3.8655 
2025-07-10 11:20:54,331 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                                            ▄▄▄▄▄▄▄▄▄▀▀▀▀▀▄▄│
│                                      ▄▄▀▀▀▗▄▄▄▄▞▀▀▀▀▀▀▀▀▀▄▄│ 0.8
│                                 ▄▄▄▄▄▄▞▀▀▀▘▄▄▄▄▞▀▀▀▀▀▀▀▘   │
│                            ▄▄▀▀▀▘ ▗▄▄▞▀▀▀▘                 │
│         ▗▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▛▀▀ ▄▄▞▀▀▘             ▄▄▄▄▄▄▞▀▀▀▀│
│       ▄▀▘             ▄▞▘▄▀▀▀         ▗▄▄▄▞▀▀▀▀▀▄▄▀▀▀▀▀▀▀▀▀│ 0.6
│       ▌            ▗▄▀▀▘       ▗▄▄▄▞▀▀▘▀▀▀▀                │
│      ▐         ▄▄▞▀▘     ▗▄▟▗▄▞▀▀                          │
│      ▐  ▐▀▀▀▗▀▀▀    ▗▄▞▀▗▛▀▀▘                              │
│     ▗▌▄▄▄▄▄▞▘    ▗▄▀▗▄▄▞▘                                  │ 0.4
│  ▄▞▀▗▘ ▗▘     ▗▄▀▘▗▞▘                                      │
│▄▀   ▐  ▐   ▗▄▞▘ ▗▞▘                                        │
│     ▌  ▌▐▀▄▄▄▄▞▀▘                                          │
│   ▗▗▘▀▗▗▘                                                  │ 0.2
│▄▀▀▗▘  ▗▘                                                   │
│  ▗▘ ▗▞▘                                                    │
│▀▀▘▀▀▘                                                      │
└────────────────────────────────────────────────────────────┘
          5          10          15          20          25
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:20:54,331 - INFO - machamp.model.callback - Performance of 3.8655 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_26.pt
2025-07-10 11:20:56,166 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_25.pt
2025-07-10 11:20:56,291 - INFO - machamp.model.trainer - 
2025-07-10 11:20:56,291 - INFO - machamp.model.trainer - Epoch 27/60: training
2025-07-10 11:21:22,827 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:21:22,828 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.6105721393034826e-05
2025-07-10 11:21:22,828 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.6105721393034826e-05
2025-07-10 11:21:22,828 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 6.120174129353234e-06
2025-07-10 11:21:22,828 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:21:22,828 - INFO - machamp.model.trainer - Epoch 27: evaluating on dev
2025-07-10 11:21:24,360 - INFO - machamp.model.callback - epoch       : 27/60
2025-07-10 11:21:24,360 - INFO - machamp.model.callback - best_epoch  : 27
2025-07-10 11:21:24,360 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:21:24,360 - INFO - machamp.model.callback - cur_ram     : 2.1407
2025-07-10 11:21:24,360 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:21:24,360 - INFO - machamp.model.callback - time_total  : 0:13:15
2025-07-10 11:21:24,360 - INFO - machamp.model.callback - 
2025-07-10 11:21:24,360 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:21:24,360 - INFO - machamp.model.callback - Best (27)                                                  
2025-07-10 11:21:24,360 - INFO - machamp.model.callback - dependency_las     0.6469   1.8213       0.7473     0.6199 
2025-07-10 11:21:24,360 - INFO - machamp.model.callback - lemma_accuracy     0.6583   0.7230       0.8036     0.8105 
2025-07-10 11:21:24,360 - INFO - machamp.model.callback - morph_accuracy     1.3079   1.3227       0.7204     0.6887 
2025-07-10 11:21:24,360 - INFO - machamp.model.callback - upos_accuracy      0.1543   0.1692       0.9159     0.8967 
2025-07-10 11:21:24,360 - INFO - machamp.model.callback - xpos_accuracy      0.2157   0.2207       0.8809     0.8608 
2025-07-10 11:21:24,360 - INFO - machamp.model.callback - sum                2.9832   4.2568       4.0681     3.8766 
2025-07-10 11:21:24,360 - INFO - machamp.model.callback - Epoch 27                                                   
2025-07-10 11:21:24,361 - INFO - machamp.model.callback - dependency_las     0.6469   1.8213       0.7473     0.6199 
2025-07-10 11:21:24,361 - INFO - machamp.model.callback - lemma_accuracy     0.6583   0.7230       0.8036     0.8105 
2025-07-10 11:21:24,361 - INFO - machamp.model.callback - morph_accuracy     1.3079   1.3227       0.7204     0.6887 
2025-07-10 11:21:24,361 - INFO - machamp.model.callback - upos_accuracy      0.1543   0.1692       0.9159     0.8967 
2025-07-10 11:21:24,361 - INFO - machamp.model.callback - xpos_accuracy      0.2157   0.2207       0.8809     0.8608 
2025-07-10 11:21:24,361 - INFO - machamp.model.callback - sum                2.9832   4.2568       4.0681     3.8766 
2025-07-10 11:21:24,369 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                                           ▗▄▄▄▄▄▄▄▄▞▀▀▀▀▄▄▄│
│                                    ▗▄▄▀▀▀▗▄▄▄▄▄▀▀▀▀▀▀▀▀▀▘▄▄│ 0.8
│                                ▗▄▄▄▄▄▄▀▀▀▘▄▄▄▄▞▀▀▀▀▀▀▀▘    │
│                           ▄▄▀▀▀▀ ▗▄▄▞▀▀▀▘                  │
│         ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▞▀▀▘▄▄▄▀▀▀            ▗▄▄▄▄▄▄▀▀▀▀▀▀│
│       ▄▀▘            ▗▞▘▄▀▀▀         ▗▄▄▄▞▀▀▀▀▘▄▄▀▀▄▄▄▞▀▀▀▀│ 0.6
│      ▗▘           ▗▄▀▘▘       ▗▄▄▄▞▀▀▘▀▀▀▘                 │
│      ▐        ▗▄▄▀▘     ▗▄▄▗▄▞▀▀                           │
│      ▌  ▛▀▀▀▞▀▘▘   ▗▄▄▀▗▞▀▀▘                               │
│     ▗▘▄▄▄▄▄▀     ▄▞▘▄▄▞▘                                   │ 0.4
│  ▗▞▀▐▘ ▐      ▄▞▀ ▄▀                                       │
│▄▀▘  ▞  ▌   ▗▄▀  ▄▀                                         │
│     ▌ ▗▘▛▄▄▄▄▄▀▀                                           │
│   ▗▐▞▀▐▞                                                   │ 0.2
│▄▀▀▗▘  ▞                                                    │
│  ▗▘ ▄▞▘                                                    │
│▀▀▘▀▀▘                                                      │
└────────────────────────────────────────────────────────────┘
          5         10          15         20          25
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:21:24,369 - INFO - machamp.model.callback - Performance of 3.8766 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_27.pt
2025-07-10 11:21:26,123 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_26.pt
2025-07-10 11:21:26,250 - INFO - machamp.model.trainer - 
2025-07-10 11:21:26,250 - INFO - machamp.model.trainer - Epoch 28/60: training
2025-07-10 11:21:52,902 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:21:52,903 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.5636608623548922e-05
2025-07-10 11:21:52,903 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.5636608623548922e-05
2025-07-10 11:21:52,903 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 5.941911276948591e-06
2025-07-10 11:21:52,903 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:21:52,903 - INFO - machamp.model.trainer - Epoch 28: evaluating on dev
2025-07-10 11:21:54,480 - INFO - machamp.model.callback - epoch       : 28/60
2025-07-10 11:21:54,480 - INFO - machamp.model.callback - best_epoch  : 28
2025-07-10 11:21:54,480 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:21:54,480 - INFO - machamp.model.callback - cur_ram     : 2.1278
2025-07-10 11:21:54,480 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:21:54,480 - INFO - machamp.model.callback - time_total  : 0:13:45
2025-07-10 11:21:54,480 - INFO - machamp.model.callback - 
2025-07-10 11:21:54,480 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:21:54,480 - INFO - machamp.model.callback - Best (28)                                                  
2025-07-10 11:21:54,480 - INFO - machamp.model.callback - dependency_las     0.6158   1.8075       0.7523     0.6233 
2025-07-10 11:21:54,480 - INFO - machamp.model.callback - lemma_accuracy     0.6411   0.7082       0.8108     0.8155 
2025-07-10 11:21:54,480 - INFO - machamp.model.callback - morph_accuracy     1.2598   1.2862       0.7286     0.6978 
2025-07-10 11:21:54,481 - INFO - machamp.model.callback - upos_accuracy      0.1420   0.1617       0.9212     0.8963 
2025-07-10 11:21:54,481 - INFO - machamp.model.callback - xpos_accuracy      0.2025   0.2140       0.8901     0.8604 
2025-07-10 11:21:54,481 - INFO - machamp.model.callback - sum                2.8612   4.1775       4.1029     3.8933 
2025-07-10 11:21:54,481 - INFO - machamp.model.callback - Epoch 28                                                   
2025-07-10 11:21:54,481 - INFO - machamp.model.callback - dependency_las     0.6158   1.8075       0.7523     0.6233 
2025-07-10 11:21:54,481 - INFO - machamp.model.callback - lemma_accuracy     0.6411   0.7082       0.8108     0.8155 
2025-07-10 11:21:54,481 - INFO - machamp.model.callback - morph_accuracy     1.2598   1.2862       0.7286     0.6978 
2025-07-10 11:21:54,481 - INFO - machamp.model.callback - upos_accuracy      0.1420   0.1617       0.9212     0.8963 
2025-07-10 11:21:54,481 - INFO - machamp.model.callback - xpos_accuracy      0.2025   0.2140       0.8901     0.8604 
2025-07-10 11:21:54,481 - INFO - machamp.model.callback - sum                2.8612   4.1775       4.1029     3.8933 
2025-07-10 11:21:54,487 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                                          ▄▄▄▄▄▄▄▄▞▀▀▀▀▄▄▄▄▄│
│                                   ▄▄▞▀▀▀▄▄▄▄▄▛▀▀▀▀▀▀▀▀▄▄▄▄▄│ 0.8
│                               ▗▄▄▄▄▄▞▀▀▀▄▄▄▄▄▀▀▀▀▀▀▀▘      │
│                          ▄▄▀▀▀▘ ▗▄▄▀▀▀▀                   ▄│
│         ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▟▀▀▘▄▄▞▀▀▘            ▄▄▄▄▄▄▀▀▀▀▀▀▀ │
│      ▗▞▀            ▗▞▘▄▀▀▀         ▄▄▄▄▀▀▀▀▀▄▄▀▀▄▄▄▛▀▀▀▀▀▀│ 0.6
│      ▐            ▄▞▘▘       ▗▗▄▄▀▀▀▀▀▀▀                   │
│      ▌        ▄▄▞▀     ▗▄▄▗▄▞▀▘                            │
│      ▌ ▗▀▀▀▗▀▀▀    ▄▄▀▗▟▀▀▘                                │
│     ▐▄▄▄▄▄▞▘    ▗▞▀▄▄▞▘                                    │ 0.4
│  ▄▞▀▞  ▌     ▗▄▀▘▄▀▘                                       │
│▄▀   ▌ ▗▘   ▄▞▘ ▗▞                                          │
│    ▗▘ ▐▗▀▄▄▄▄▞▀▘                                           │
│   ▗▞▞▀▌▌                                                   │ 0.2
│▄▀▀▐  ▗▞                                                    │
│  ▗▘ ▄▀                                                     │
│▀▀▘▀▀▘                                                      │
└────────────────────────────────────────────────────────────┘
         5          10         15         20         25
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:21:54,487 - INFO - machamp.model.callback - Performance of 3.8933 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_28.pt
2025-07-10 11:21:56,266 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_27.pt
2025-07-10 11:21:56,395 - INFO - machamp.model.trainer - 
2025-07-10 11:21:56,395 - INFO - machamp.model.trainer - Epoch 29/60: training
2025-07-10 11:22:22,940 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:22:22,940 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.5167495854063019e-05
2025-07-10 11:22:22,940 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.5167495854063019e-05
2025-07-10 11:22:22,940 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 5.763648424543947e-06
2025-07-10 11:22:22,940 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:22:22,940 - INFO - machamp.model.trainer - Epoch 29: evaluating on dev
2025-07-10 11:22:24,499 - INFO - machamp.model.callback - epoch       : 29/60
2025-07-10 11:22:24,499 - INFO - machamp.model.callback - best_epoch  : 29
2025-07-10 11:22:24,499 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:22:24,499 - INFO - machamp.model.callback - cur_ram     : 2.1407
2025-07-10 11:22:24,499 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:22:24,499 - INFO - machamp.model.callback - time_total  : 0:14:15
2025-07-10 11:22:24,499 - INFO - machamp.model.callback - 
2025-07-10 11:22:24,499 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:22:24,499 - INFO - machamp.model.callback - Best (29)                                                  
2025-07-10 11:22:24,499 - INFO - machamp.model.callback - dependency_las     0.5541   1.8501       0.7725     0.6169 
2025-07-10 11:22:24,499 - INFO - machamp.model.callback - lemma_accuracy     0.6196   0.6975       0.8169     0.8165 
2025-07-10 11:22:24,499 - INFO - machamp.model.callback - morph_accuracy     1.2178   1.2513       0.7353     0.7008 
2025-07-10 11:22:24,499 - INFO - machamp.model.callback - upos_accuracy      0.1322   0.1588       0.9271     0.8987 
2025-07-10 11:22:24,499 - INFO - machamp.model.callback - xpos_accuracy      0.1939   0.2080       0.8943     0.8641 
2025-07-10 11:22:24,499 - INFO - machamp.model.callback - sum                2.7176   4.1657       4.1461     3.8970 
2025-07-10 11:22:24,499 - INFO - machamp.model.callback - Epoch 29                                                   
2025-07-10 11:22:24,499 - INFO - machamp.model.callback - dependency_las     0.5541   1.8501       0.7725     0.6169 
2025-07-10 11:22:24,499 - INFO - machamp.model.callback - lemma_accuracy     0.6196   0.6975       0.8169     0.8165 
2025-07-10 11:22:24,500 - INFO - machamp.model.callback - morph_accuracy     1.2178   1.2513       0.7353     0.7008 
2025-07-10 11:22:24,500 - INFO - machamp.model.callback - upos_accuracy      0.1322   0.1588       0.9271     0.8987 
2025-07-10 11:22:24,500 - INFO - machamp.model.callback - xpos_accuracy      0.1939   0.2080       0.8943     0.8641 
2025-07-10 11:22:24,500 - INFO - machamp.model.callback - sum                2.7176   4.1657       4.1461     3.8970 
2025-07-10 11:22:24,508 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                                         ▄▄▄▄▄▄▄▄▞▀▀▀▀▗▄▄▄▄▄│
│                                  ▄▄▞▀▀▀▄▄▄▄▄▀▀▀▀▀▀▀▀▀▘▄▄▄▄▄│ 0.8
│                              ▗▄▄▄▄▄▀▀▀▀▄▄▄▄▞▀▀▀▀▀▀▀▘       │
│                         ▗▄▀▀▀▀ ▗▄▄▀▀▀▀                  ▗▄▄│
│        ▗▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▀▘▘▄▄▞▀▀▘           ▗▄▄▄▄▄▞▀▀▀▀▀▀▘  │
│      ▗▞▘            ▄▀▄▞▀▀         ▄▄▄▄▀▀▀▀▘▄▞▚▄▄▄▞▀▀▀▀▀▀▀▀│ 0.6
│      ▞           ▗▞▀▀       ▗▗▄▄▀▀▀▀▀▀▀                    │
│      ▌       ▗▄▄▀▘     ▄▄▗▄▞▀▘                             │
│     ▐  ▐▀▀▀▄▀▘▘   ▄▄▞▀▄▀▀▘                                 │
│     ▐▄▄▄▄▄▞     ▄▀▗▄▄▀                                     │ 0.4
│  ▄▞▀▌ ▗▘     ▄▞▀▗▞▘                                        │
│▄▀   ▌ ▐   ▗▄▀  ▞▘                                          │
│    ▐  ▞▐▚▄▄▄▄▀▀                                            │
│   ▗▞▀▀▗▘                                                   │ 0.2
│▄▀▀▞  ▗▘                                                    │
│  ▄▘▗▄▀                                                     │
│▀▀▀▀▀▘                                                      │
└────────────────────────────────────────────────────────────┘
         5         10         15        20         25
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:22:24,508 - INFO - machamp.model.callback - Performance of 3.8970 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_29.pt
2025-07-10 11:22:26,272 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_28.pt
2025-07-10 11:22:26,400 - INFO - machamp.model.trainer - 
2025-07-10 11:22:26,400 - INFO - machamp.model.trainer - Epoch 30/60: training
2025-07-10 11:22:52,939 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:22:52,939 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.4698383084577117e-05
2025-07-10 11:22:52,939 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.4698383084577117e-05
2025-07-10 11:22:52,939 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 5.585385572139304e-06
2025-07-10 11:22:52,939 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:22:52,939 - INFO - machamp.model.trainer - Epoch 30: evaluating on dev
2025-07-10 11:22:54,479 - INFO - machamp.model.callback - epoch       : 30/60
2025-07-10 11:22:54,479 - INFO - machamp.model.callback - best_epoch  : 30
2025-07-10 11:22:54,479 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:22:54,479 - INFO - machamp.model.callback - cur_ram     : 2.1408
2025-07-10 11:22:54,479 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:22:54,479 - INFO - machamp.model.callback - time_total  : 0:14:45
2025-07-10 11:22:54,479 - INFO - machamp.model.callback - 
2025-07-10 11:22:54,479 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:22:54,479 - INFO - machamp.model.callback - Best (30)                                                  
2025-07-10 11:22:54,479 - INFO - machamp.model.callback - dependency_las     0.5399   1.8371       0.7778     0.6260 
2025-07-10 11:22:54,479 - INFO - machamp.model.callback - lemma_accuracy     0.6061   0.6882       0.8194     0.8212 
2025-07-10 11:22:54,479 - INFO - machamp.model.callback - morph_accuracy     1.1768   1.2266       0.7407     0.7071 
2025-07-10 11:22:54,480 - INFO - machamp.model.callback - upos_accuracy      0.1254   0.1573       0.9320     0.9010 
2025-07-10 11:22:54,480 - INFO - machamp.model.callback - xpos_accuracy      0.1849   0.2054       0.8974     0.8695 
2025-07-10 11:22:54,480 - INFO - machamp.model.callback - sum                2.6331   4.1145       4.1673     3.9249 
2025-07-10 11:22:54,480 - INFO - machamp.model.callback - Epoch 30                                                   
2025-07-10 11:22:54,480 - INFO - machamp.model.callback - dependency_las     0.5399   1.8371       0.7778     0.6260 
2025-07-10 11:22:54,480 - INFO - machamp.model.callback - lemma_accuracy     0.6061   0.6882       0.8194     0.8212 
2025-07-10 11:22:54,480 - INFO - machamp.model.callback - morph_accuracy     1.1768   1.2266       0.7407     0.7071 
2025-07-10 11:22:54,480 - INFO - machamp.model.callback - upos_accuracy      0.1254   0.1573       0.9320     0.9010 
2025-07-10 11:22:54,480 - INFO - machamp.model.callback - xpos_accuracy      0.1849   0.2054       0.8974     0.8695 
2025-07-10 11:22:54,480 - INFO - machamp.model.callback - sum                2.6331   4.1145       4.1673     3.9249 
2025-07-10 11:22:54,488 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                                        ▄▄▄▄▄▄▄▄▄▀▀▀▀▀▀▀▄▄▄▄│
│                                 ▄▄▞▀▀▀▄▄▄▄▄▀▀▀▀▀▀▀▀▀▀▀▀▄▄▄▄│ 0.8
│                             ▗▗▄▄▄▄▛▀▀▀▄▄▄▄▞▀▀▀▀▀▀▀▀        │
│                        ▗▄▞▀▀▀▘▗▄▄▀▀▀▀                  ▄▄▄▄│
│        ▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▀▘▀▗▄▟▀▀▘           ▗▄▄▄▄▄▀▀▀▀▀▀▀    │
│      ▄▞▘           ▄▀▄▄▀▀▘        ▄▄▄▄▀▀▀▀▘▄▄▄▄▄▄▀▀▀▀▀▀▀▜▀▀│ 0.6
│      ▌          ▗▄▀▀▘      ▗▄▄▄▀▀▀▀▀▀▘                     │
│     ▗▘       ▄▄▞▀     ▄▄▞▄▞▀▀                              │
│     ▐  ▛▀▀▗▞▀▀   ▗▄▄▀▄▀▀▀                                  │
│     ▞▄▄▄▄▄▘    ▗▞▘▄▄▀                                      │ 0.4
│  ▄▞▀▌ ▐      ▄▀▘▄▀                                         │
│▄▀  ▐  ▞   ▄▞▀ ▄▀                                           │
│    ▐  ▌▛▄▄▄▄▞▀                                             │
│   ▗▌▀▐▐                                                    │ 0.2
│▄▀▀▞  ▗▘                                                    │
│  ▟ ▗▞▘                                                     │
│▀▀▀▀▘                                                       │
└────────────────────────────────────────────────────────────┘
                  10                   20                  30
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:22:54,489 - INFO - machamp.model.callback - Performance of 3.9249 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_30.pt
2025-07-10 11:22:56,261 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_29.pt
2025-07-10 11:22:56,389 - INFO - machamp.model.trainer - 
2025-07-10 11:22:56,389 - INFO - machamp.model.trainer - Epoch 31/60: training
2025-07-10 11:23:22,975 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:23:22,975 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.422927031509121e-05
2025-07-10 11:23:22,975 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.422927031509121e-05
2025-07-10 11:23:22,975 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 5.40712271973466e-06
2025-07-10 11:23:22,975 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:23:22,976 - INFO - machamp.model.trainer - Epoch 31: evaluating on dev
2025-07-10 11:23:24,532 - INFO - machamp.model.callback - epoch       : 31/60
2025-07-10 11:23:24,532 - INFO - machamp.model.callback - best_epoch  : 31
2025-07-10 11:23:24,532 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:23:24,532 - INFO - machamp.model.callback - cur_ram     : 2.1408
2025-07-10 11:23:24,532 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:23:24,532 - INFO - machamp.model.callback - time_total  : 0:15:15
2025-07-10 11:23:24,532 - INFO - machamp.model.callback - 
2025-07-10 11:23:24,533 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:23:24,533 - INFO - machamp.model.callback - Best (31)                                                  
2025-07-10 11:23:24,533 - INFO - machamp.model.callback - dependency_las     0.4864   1.8774       0.7882     0.6286 
2025-07-10 11:23:24,533 - INFO - machamp.model.callback - lemma_accuracy     0.5893   0.6782       0.8242     0.8256 
2025-07-10 11:23:24,533 - INFO - machamp.model.callback - morph_accuracy     1.1313   1.1990       0.7510     0.7128 
2025-07-10 11:23:24,533 - INFO - machamp.model.callback - upos_accuracy      0.1163   0.1530       0.9366     0.9031 
2025-07-10 11:23:24,533 - INFO - machamp.model.callback - xpos_accuracy      0.1755   0.1991       0.9049     0.8739 
2025-07-10 11:23:24,533 - INFO - machamp.model.callback - sum                2.4990   4.1068       4.2049     3.9440 
2025-07-10 11:23:24,533 - INFO - machamp.model.callback - Epoch 31                                                   
2025-07-10 11:23:24,533 - INFO - machamp.model.callback - dependency_las     0.4864   1.8774       0.7882     0.6286 
2025-07-10 11:23:24,533 - INFO - machamp.model.callback - lemma_accuracy     0.5893   0.6782       0.8242     0.8256 
2025-07-10 11:23:24,533 - INFO - machamp.model.callback - morph_accuracy     1.1313   1.1990       0.7510     0.7128 
2025-07-10 11:23:24,533 - INFO - machamp.model.callback - upos_accuracy      0.1163   0.1530       0.9366     0.9031 
2025-07-10 11:23:24,533 - INFO - machamp.model.callback - xpos_accuracy      0.1755   0.1991       0.9049     0.8739 
2025-07-10 11:23:24,533 - INFO - machamp.model.callback - sum                2.4990   4.1068       4.2049     3.9440 
2025-07-10 11:23:24,542 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                                       ▄▄▄▄▄▄▄▄▄▞▀▀▀▀▀▀▄▄▄▄▄│
│                                ▄▄▞▀▀▀▄▄▄▄▄▀▀▀▀▀▀▀▀▀▀▀▀▄▄▄▄▄│ 0.8
│                             ▄▄▄▄▄▞▀▀▀▄▄▄▄▄▀▀▀▀▀▀▀▀         │
│                        ▄▞▀▀▀▀▗▄▄▀▀▀▀                  ▗▄▄▄▄│
│                  ▗▄▄▄▞▀▀▗▄▄▛▀▘           ▗▄▄▄▄▞▀▀▀▀▀▀▀▘    │
│      ▄▀▀▀▀▀▀▀▀▀▀▀▘▗▞▘▄▀▀▘       ▗▄▄▄▟▀▀▀▀▘▄▄▄▄▄▞▀▀▀▀▀▀▄▀▀▀▀│ 0.6
│      ▌          ▗▞▘▀       ▄▄▄▀▀▀▀▀▀▘                      │
│     ▐        ▄▄▀▘    ▗▄▄▄▞▀▀                               │
│     ▞  ▛▀▀▗▀▀▘   ▄▄▞▗▞▀▀                                   │
│     ▌▗▄▄▄▞▘    ▄▀▗▄▞▘                                      │ 0.4
│  ▄▞▗▘▘▞     ▗▞▀▗▞▘                                         │
│▄▀  ▐  ▌   ▄▀▘ ▞▘                                           │
│    ▞ ▗▘▙▄▄▄▄▀▀                                             │
│   ▄▌▀▐▞                                                    │ 0.2
│▄▀▀▞  ▐                                                     │
│  ▞ ▗▞▘                                                     │
│▀▀▀▀▘                                                       │
└────────────────────────────────────────────────────────────┘
                  10                 20                  30
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:23:24,542 - INFO - machamp.model.callback - Performance of 3.9440 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_31.pt
2025-07-10 11:23:26,277 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_30.pt
2025-07-10 11:23:26,411 - INFO - machamp.model.trainer - 
2025-07-10 11:23:26,411 - INFO - machamp.model.trainer - Epoch 32/60: training
2025-07-10 11:23:53,006 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:23:53,006 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.3760157545605309e-05
2025-07-10 11:23:53,006 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.3760157545605309e-05
2025-07-10 11:23:53,006 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 5.228859867330017e-06
2025-07-10 11:23:53,006 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:23:53,007 - INFO - machamp.model.trainer - Epoch 32: evaluating on dev
2025-07-10 11:23:54,659 - INFO - machamp.model.callback - epoch       : 32/60
2025-07-10 11:23:54,659 - INFO - machamp.model.callback - best_epoch  : 32
2025-07-10 11:23:54,659 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:23:54,659 - INFO - machamp.model.callback - cur_ram     : 2.1408
2025-07-10 11:23:54,659 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:23:54,659 - INFO - machamp.model.callback - time_total  : 0:15:45
2025-07-10 11:23:54,659 - INFO - machamp.model.callback - 
2025-07-10 11:23:54,659 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:23:54,659 - INFO - machamp.model.callback - Best (32)                                                  
2025-07-10 11:23:54,659 - INFO - machamp.model.callback - dependency_las     0.4495   1.8865       0.8002     0.6233 
2025-07-10 11:23:54,659 - INFO - machamp.model.callback - lemma_accuracy     0.5739   0.6676       0.8294     0.8266 
2025-07-10 11:23:54,659 - INFO - machamp.model.callback - morph_accuracy     1.0975   1.1804       0.7584     0.7159 
2025-07-10 11:23:54,660 - INFO - machamp.model.callback - upos_accuracy      0.1124   0.1510       0.9367     0.9084 
2025-07-10 11:23:54,660 - INFO - machamp.model.callback - xpos_accuracy      0.1686   0.1973       0.9085     0.8786 
2025-07-10 11:23:54,660 - INFO - machamp.model.callback - sum                2.4019   4.0827       4.2333     3.9527 
2025-07-10 11:23:54,660 - INFO - machamp.model.callback - Epoch 32                                                   
2025-07-10 11:23:54,660 - INFO - machamp.model.callback - dependency_las     0.4495   1.8865       0.8002     0.6233 
2025-07-10 11:23:54,660 - INFO - machamp.model.callback - lemma_accuracy     0.5739   0.6676       0.8294     0.8266 
2025-07-10 11:23:54,660 - INFO - machamp.model.callback - morph_accuracy     1.0975   1.1804       0.7584     0.7159 
2025-07-10 11:23:54,660 - INFO - machamp.model.callback - upos_accuracy      0.1124   0.1510       0.9367     0.9084 
2025-07-10 11:23:54,660 - INFO - machamp.model.callback - xpos_accuracy      0.1686   0.1973       0.9085     0.8786 
2025-07-10 11:23:54,660 - INFO - machamp.model.callback - sum                2.4019   4.0827       4.2333     3.9527 
2025-07-10 11:23:54,669 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                                      ▗▄▄▄▄▄▄▄▄▄▄▞▀▀▀▀▀▗▄▄▄▄│
│                               ▗▄▄▀▀▀▗▄▄▄▄▄▀▀▀▀▀▀▀▀▀▀▀▀▘▄▄▄▄│
│                            ▗▄▗▄▄▞▀▀▀▘▄▄▄▄▄▀▀▀▀▀▀▀▀▘        │ 0.8
│                       ▗▄▀▀▀▀▀▘▄▞▀▀▀▘                  ▄▄▄▄▄│
│                   ▄▄▄▀▘▀▄▄▞▀▀            ▄▄▄▄▄▀▀▀▀▀▀▀▀     │
│      ▄▀▀▀▀▀▀▀▀▀▀▀▀▄▀▄▄▀▀        ▄▄▄▄▀▀▀▀▀▄▄▄▄▄▞▀▚▄▞▚▄▄▀▀▀▀▀│ 0.6
│     ▐          ▗▄▀▀▘      ▗▗▄▞▀▀▀▀▀▀                       │
│     ▐       ▗▄▟▀     ▄▄▗▄▀▀▘                               │
│     ▌ ▐▀▀▀▞▀▘    ▄▄▗▄▀▀▘                                   │
│     ▌ ▐▄▄▞    ▗▞▀▄▄▘                                       │ 0.4
│  ▄▞▐▀▀▀     ▄▀▘▄▀                                          │
│▄▀  ▐ ▗▘  ▗▄▀ ▗▀                                            │
│    ▌ ▐▗▄▄▄▄▄▀▘                                             │
│   ▗▘▀▞▌                                                    │ 0.2
│▄▀▀▞  ▞                                                     │
│  ▞ ▗▞▘                                                     │
│▀▀▀▀▘                                                       │
└────────────────────────────────────────────────────────────┘
                 10                 20                  30
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:23:54,669 - INFO - machamp.model.callback - Performance of 3.9527 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_32.pt
2025-07-10 11:23:56,415 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_31.pt
2025-07-10 11:23:56,545 - INFO - machamp.model.trainer - 
2025-07-10 11:23:56,545 - INFO - machamp.model.trainer - Epoch 33/60: training
2025-07-10 11:24:23,134 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:24:23,134 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.3291044776119404e-05
2025-07-10 11:24:23,134 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.3291044776119404e-05
2025-07-10 11:24:23,134 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 5.050597014925373e-06
2025-07-10 11:24:23,134 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:24:23,135 - INFO - machamp.model.trainer - Epoch 33: evaluating on dev
2025-07-10 11:24:24,698 - INFO - machamp.model.callback - epoch       : 33/60
2025-07-10 11:24:24,699 - INFO - machamp.model.callback - best_epoch  : 33
2025-07-10 11:24:24,699 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:24:24,699 - INFO - machamp.model.callback - cur_ram     : 2.1279
2025-07-10 11:24:24,699 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:24:24,699 - INFO - machamp.model.callback - time_total  : 0:16:15
2025-07-10 11:24:24,699 - INFO - machamp.model.callback - 
2025-07-10 11:24:24,699 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:24:24,699 - INFO - machamp.model.callback - Best (33)                                                  
2025-07-10 11:24:24,699 - INFO - machamp.model.callback - dependency_las     0.4231   1.8988       0.8079     0.6266 
2025-07-10 11:24:24,699 - INFO - machamp.model.callback - lemma_accuracy     0.5593   0.6605       0.8301     0.8282 
2025-07-10 11:24:24,699 - INFO - machamp.model.callback - morph_accuracy     1.0599   1.1567       0.7628     0.7263 
2025-07-10 11:24:24,699 - INFO - machamp.model.callback - upos_accuracy      0.1058   0.1501       0.9426     0.9088 
2025-07-10 11:24:24,699 - INFO - machamp.model.callback - xpos_accuracy      0.1635   0.1950       0.9129     0.8833 
2025-07-10 11:24:24,699 - INFO - machamp.model.callback - sum                2.3118   4.0613       4.2562     3.9732 
2025-07-10 11:24:24,699 - INFO - machamp.model.callback - Epoch 33                                                   
2025-07-10 11:24:24,699 - INFO - machamp.model.callback - dependency_las     0.4231   1.8988       0.8079     0.6266 
2025-07-10 11:24:24,699 - INFO - machamp.model.callback - lemma_accuracy     0.5593   0.6605       0.8301     0.8282 
2025-07-10 11:24:24,699 - INFO - machamp.model.callback - morph_accuracy     1.0599   1.1567       0.7628     0.7263 
2025-07-10 11:24:24,699 - INFO - machamp.model.callback - upos_accuracy      0.1058   0.1501       0.9426     0.9088 
2025-07-10 11:24:24,699 - INFO - machamp.model.callback - xpos_accuracy      0.1635   0.1950       0.9129     0.8833 
2025-07-10 11:24:24,700 - INFO - machamp.model.callback - sum                2.3118   4.0613       4.2562     3.9732 
2025-07-10 11:24:24,708 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                                     ▄▄▄▄▄▄▄▄▄▄▄▞▀▀▀▀▀▄▄▄▄▄▄│
│                              ▗▄▄▀▀▀▗▄▄▄▄▞▀▀▀▀▀▀▀▀▀▀▀▀▄▄▄▄▄▄│
│                           ▗▄▗▄▄▞▀▀▀▘▄▄▄▄▞▀▀▀▀▀▀▀▀          │ 0.8
│                      ▗▄▛▀▀▀▀▘▄▞▀▀▀▘                 ▗▄▄▄▄▄▄│
│                  ▗▄▄▞▀▀▄▄▞▀▀            ▄▄▄▄▞▀▀▀▀▀▀▀▘      │
│     ▗▄▀▀▀▀▀▀▀▀▀▀▀▘▞▘▄▀▀        ▄▄▄▄▀▀▀▀▀▄▄▄▄▄▀▀▄▄▚▄▄▞▀▀▀▀▀▀│ 0.6
│     ▐          ▗▞▘▀      ▗▗▄▞▀▀▀▀▀▀                        │
│     ▞       ▄▄▛▘    ▗▄▗▄▞▀▘                                │
│     ▌ ▐▀▀▗▛▀▀   ▄▄▞▄▞▀▘                                    │
│    ▗▘ ▌▄▄▘    ▄▀▗▄▞                                        │ 0.4
│  ▄▀▐▀▀▀    ▗▞▀▗▞▘                                          │
│▄▀  ▞ ▐   ▄▞▘ ▞▘                                            │
│    ▌ ▞▗▄▄▄▄▞▀                                              │
│   ▗▘▀▗▘                                                    │ 0.2
│▄▀▀▌ ▗▞                                                     │
│  ▞ ▄▀                                                      │
│▀▀▀▀▘                                                       │
└────────────────────────────────────────────────────────────┘
                10                 20                 30
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:24:24,708 - INFO - machamp.model.callback - Performance of 3.9732 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_33.pt
2025-07-10 11:24:26,493 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_32.pt
2025-07-10 11:24:26,632 - INFO - machamp.model.trainer - 
2025-07-10 11:24:26,632 - INFO - machamp.model.trainer - Epoch 34/60: training
2025-07-10 11:24:53,249 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:24:53,249 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.2821932006633499e-05
2025-07-10 11:24:53,249 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.2821932006633499e-05
2025-07-10 11:24:53,249 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 4.87233416252073e-06
2025-07-10 11:24:53,249 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:24:53,249 - INFO - machamp.model.trainer - Epoch 34: evaluating on dev
2025-07-10 11:24:54,805 - INFO - machamp.model.callback - epoch       : 34/60
2025-07-10 11:24:54,805 - INFO - machamp.model.callback - best_epoch  : 34
2025-07-10 11:24:54,805 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:24:54,805 - INFO - machamp.model.callback - cur_ram     : 2.1408
2025-07-10 11:24:54,805 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:24:54,805 - INFO - machamp.model.callback - time_total  : 0:16:45
2025-07-10 11:24:54,805 - INFO - machamp.model.callback - 
2025-07-10 11:24:54,805 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:24:54,805 - INFO - machamp.model.callback - Best (34)                                                  
2025-07-10 11:24:54,805 - INFO - machamp.model.callback - dependency_las     0.3956   1.9429       0.8178     0.6293 
2025-07-10 11:24:54,805 - INFO - machamp.model.callback - lemma_accuracy     0.5492   0.6529       0.8349     0.8313 
2025-07-10 11:24:54,805 - INFO - machamp.model.callback - morph_accuracy     1.0287   1.1334       0.7691     0.7300 
2025-07-10 11:24:54,805 - INFO - machamp.model.callback - upos_accuracy      0.0989   0.1473       0.9464     0.9094 
2025-07-10 11:24:54,806 - INFO - machamp.model.callback - xpos_accuracy      0.1559   0.1920       0.9154     0.8833 
2025-07-10 11:24:54,806 - INFO - machamp.model.callback - sum                2.2283   4.0684       4.2837     3.9832 
2025-07-10 11:24:54,806 - INFO - machamp.model.callback - Epoch 34                                                   
2025-07-10 11:24:54,806 - INFO - machamp.model.callback - dependency_las     0.3956   1.9429       0.8178     0.6293 
2025-07-10 11:24:54,806 - INFO - machamp.model.callback - lemma_accuracy     0.5492   0.6529       0.8349     0.8313 
2025-07-10 11:24:54,806 - INFO - machamp.model.callback - morph_accuracy     1.0287   1.1334       0.7691     0.7300 
2025-07-10 11:24:54,806 - INFO - machamp.model.callback - upos_accuracy      0.0989   0.1473       0.9464     0.9094 
2025-07-10 11:24:54,806 - INFO - machamp.model.callback - xpos_accuracy      0.1559   0.1920       0.9154     0.8833 
2025-07-10 11:24:54,806 - INFO - machamp.model.callback - sum                2.2283   4.0684       4.2837     3.9832 
2025-07-10 11:24:54,814 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                                    ▄▄▄▄▄▄▄▄▄▄▄▚▄▀▀▀▄▄▄▄▄▄▄▄│
│                             ▗▄▄▀▀▀▄▄▄▄▄▞▀▀▀▀▀▀▀▀▀▀▀▘▄▄▄▄▄▄▄│
│                          ▗▄▀▄▄▄▀▀▀▄▄▄▄▄▞▀▀▀▀▀▀▀▘           │ 0.8
│                      ▄▞▀▀▀▀▀▄▟▀▀▀▘                 ▄▄▄▄▄▄▄▄│
│                  ▄▄▞▀▀▘▄▄▀▀           ▗▄▄▄▄▞▀▀▀▀▀▀▀        │
│     ▗▞▀▀▀▀▀▀▀▀▀▀▀▄▘▄▞▀▘       ▗▄▄▄▀▀▀▀▘▄▄▄▄▞▀▚▄▟▄▄▄▀▀▀▀▀▀▀▀│ 0.6
│     ▐         ▗▄▀▀      ▗▗▄▞▀▀▘▀▀▀                         │
│     ▌      ▗▄▟▀    ▗▄▄▄▞▀▘                                 │
│     ▌ ▞▀▀▗▀▀▘  ▗▄▞▗▄▀▀                                     │
│    ▐  ▌▄▄▘   ▗▞▀▄▞▘                                        │ 0.4
│  ▄▀▐▀▀▀    ▄▞▘▄▀                                           │
│▄▀  ▌ ▐   ▄▛ ▗▞                                             │
│    ▌ ▌▄▄▄▄▄▀▘                                              │
│   ▐▞▀▗▘                                                    │ 0.2
│▄▀▗▘ ▗▘                                                     │
│  ▞ ▄▀                                                      │
│▀▀▀▀▘                                                       │
└────────────────────────────────────────────────────────────┘
                10                20                30
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:24:54,815 - INFO - machamp.model.callback - Performance of 3.9832 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_34.pt
2025-07-10 11:24:56,535 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_33.pt
2025-07-10 11:24:56,666 - INFO - machamp.model.trainer - 
2025-07-10 11:24:56,666 - INFO - machamp.model.trainer - Epoch 35/60: training
2025-07-10 11:25:23,297 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:25:23,297 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.2352819237147597e-05
2025-07-10 11:25:23,297 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.2352819237147597e-05
2025-07-10 11:25:23,297 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 4.6940713101160875e-06
2025-07-10 11:25:23,297 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:25:23,297 - INFO - machamp.model.trainer - Epoch 35: evaluating on dev
2025-07-10 11:25:24,850 - INFO - machamp.model.callback - epoch       : 35/60
2025-07-10 11:25:24,850 - INFO - machamp.model.callback - best_epoch  : 35
2025-07-10 11:25:24,850 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:25:24,850 - INFO - machamp.model.callback - cur_ram     : 2.1278
2025-07-10 11:25:24,850 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:25:24,850 - INFO - machamp.model.callback - time_total  : 0:17:15
2025-07-10 11:25:24,850 - INFO - machamp.model.callback - 
2025-07-10 11:25:24,850 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:25:24,850 - INFO - machamp.model.callback - Best (35)                                                  
2025-07-10 11:25:24,850 - INFO - machamp.model.callback - dependency_las     0.3776   1.9604       0.8255     0.6286 
2025-07-10 11:25:24,850 - INFO - machamp.model.callback - lemma_accuracy     0.5329   0.6442       0.8395     0.8319 
2025-07-10 11:25:24,850 - INFO - machamp.model.callback - morph_accuracy     1.0032   1.1154       0.7742     0.7326 
2025-07-10 11:25:24,850 - INFO - machamp.model.callback - upos_accuracy      0.0973   0.1451       0.9466     0.9091 
2025-07-10 11:25:24,850 - INFO - machamp.model.callback - xpos_accuracy      0.1504   0.1892       0.9167     0.8856 
2025-07-10 11:25:24,850 - INFO - machamp.model.callback - sum                2.1614   4.0543       4.3024     3.9879 
2025-07-10 11:25:24,850 - INFO - machamp.model.callback - Epoch 35                                                   
2025-07-10 11:25:24,850 - INFO - machamp.model.callback - dependency_las     0.3776   1.9604       0.8255     0.6286 
2025-07-10 11:25:24,850 - INFO - machamp.model.callback - lemma_accuracy     0.5329   0.6442       0.8395     0.8319 
2025-07-10 11:25:24,851 - INFO - machamp.model.callback - morph_accuracy     1.0032   1.1154       0.7742     0.7326 
2025-07-10 11:25:24,851 - INFO - machamp.model.callback - upos_accuracy      0.0973   0.1451       0.9466     0.9091 
2025-07-10 11:25:24,851 - INFO - machamp.model.callback - xpos_accuracy      0.1504   0.1892       0.9167     0.8856 
2025-07-10 11:25:24,851 - INFO - machamp.model.callback - sum                2.1614   4.0543       4.3024     3.9879 
2025-07-10 11:25:24,857 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                                   ▄▄▄▄▄▄▄▄▄▄▞▄▞▀▀▗▄▄▄▄▄▄▄▄▄│
│                             ▄▄▀▀▀▄▄▄▄▄▞▀▀▀▀▀▀▀▀▀▀▀▄▄▄▄▄▄▄▄▄│
│                          ▄▞▄▄▄▀▀▀▄▄▄▄▄▞▀▀▀▀▀▀▀             │ 0.8
│                     ▗▄▀▀▀▀▀▄▄▀▀▀▘                ▗▄▄▄▄▄▄▄▄▞│
│                 ▗▄▄▀▘▘▄▄▀▀▘          ▗▄▄▄▄▀▀▀▀▀▀▀▘         │
│     ▗▞▀▀▀▀▀▀▀▀▀▀▘▞▄▄▀▀       ▗▄▄▄▀▀▀▀▘▄▄▄▄▞▀▄▄▙▄▄▞▀▀▀▀▀▀▀▀▀│ 0.6
│     ▌         ▗▞▘▘      ▄▄▞▀▀▘▀▀▀                          │
│     ▌      ▗▄▞▘    ▄▄▗▄▀▀                                  │
│    ▗▘ ▛▀▀▞▀▘   ▗▄▀▄▞▀▘                                     │
│    ▐ ▗▗▄▞    ▄▀▗▄▀                                         │ 0.4
│  ▄▀▞▀▀▘    ▄▀▗▞▘                                           │
│▄▀  ▌ ▞  ▗▄▀ ▄▘                                             │
│   ▗▘ ▌▄▄▄▄▄▀                                               │
│   ▐▞▐▞▘                                                    │ 0.2
│▄▀▗▘ ▗▘                                                     │
│  ▞ ▄▀                                                      │
│▀▀▀▀▘                                                       │
└────────────────────────────────────────────────────────────┘
               10                20                30
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:25:24,857 - INFO - machamp.model.callback - Performance of 3.9879 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_35.pt
2025-07-10 11:25:26,663 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_34.pt
2025-07-10 11:25:26,795 - INFO - machamp.model.trainer - 
2025-07-10 11:25:26,796 - INFO - machamp.model.trainer - Epoch 36/60: training
2025-07-10 11:25:53,399 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:25:53,399 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.188370646766169e-05
2025-07-10 11:25:53,399 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.188370646766169e-05
2025-07-10 11:25:53,399 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 4.515808457711443e-06
2025-07-10 11:25:53,399 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:25:53,399 - INFO - machamp.model.trainer - Epoch 36: evaluating on dev
2025-07-10 11:25:54,960 - INFO - machamp.model.callback - epoch       : 36/60
2025-07-10 11:25:54,960 - INFO - machamp.model.callback - best_epoch  : 36
2025-07-10 11:25:54,960 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:25:54,960 - INFO - machamp.model.callback - cur_ram     : 2.1279
2025-07-10 11:25:54,960 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:25:54,960 - INFO - machamp.model.callback - time_total  : 0:17:45
2025-07-10 11:25:54,960 - INFO - machamp.model.callback - 
2025-07-10 11:25:54,960 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:25:54,960 - INFO - machamp.model.callback - Best (36)                                                  
2025-07-10 11:25:54,960 - INFO - machamp.model.callback - dependency_las     0.3495   1.9898       0.8305     0.6266 
2025-07-10 11:25:54,960 - INFO - machamp.model.callback - lemma_accuracy     0.5210   0.6390       0.8408     0.8380 
2025-07-10 11:25:54,960 - INFO - machamp.model.callback - morph_accuracy     0.9788   1.0952       0.7802     0.7383 
2025-07-10 11:25:54,961 - INFO - machamp.model.callback - upos_accuracy      0.0923   0.1446       0.9502     0.9088 
2025-07-10 11:25:54,961 - INFO - machamp.model.callback - xpos_accuracy      0.1450   0.1868       0.9222     0.8873 
2025-07-10 11:25:54,961 - INFO - machamp.model.callback - sum                2.0866   4.0554       4.3239     3.9990 
2025-07-10 11:25:54,961 - INFO - machamp.model.callback - Epoch 36                                                   
2025-07-10 11:25:54,961 - INFO - machamp.model.callback - dependency_las     0.3495   1.9898       0.8305     0.6266 
2025-07-10 11:25:54,961 - INFO - machamp.model.callback - lemma_accuracy     0.5210   0.6390       0.8408     0.8380 
2025-07-10 11:25:54,961 - INFO - machamp.model.callback - morph_accuracy     0.9788   1.0952       0.7802     0.7383 
2025-07-10 11:25:54,961 - INFO - machamp.model.callback - upos_accuracy      0.0923   0.1446       0.9502     0.9088 
2025-07-10 11:25:54,961 - INFO - machamp.model.callback - xpos_accuracy      0.1450   0.1868       0.9222     0.8873 
2025-07-10 11:25:54,961 - INFO - machamp.model.callback - sum                2.0866   4.0554       4.3239     3.9990 
2025-07-10 11:25:54,967 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                                  ▄▄▄▄▄▄▄▄▄▄▙▄▞▀▀▗▄▄▄▄▄▄▄▄▄▄│
│                            ▄▄▀▀▀▄▄▄▄▄▞▀▀▀▀▀▀▀▀▀▀▘▄▄▄▄▄▄▄▄▄▄│
│                         ▄▞▄▄▄▀▀▀▄▄▄▄▄▀▀▀▀▀▀▀▀              │ 0.8
│                    ▗▄▛▀▀▀▀▄▄▀▀▀▘                ▄▄▄▄▄▄▄▄▟▀▀│
│                 ▄▄▞▀▀▄▄▞▀▘          ▗▄▄▄▄▀▀▀▀▀▀▀           │
│     ▄▞▀▀▀▀▀▀▀▀▀▀▄▘▄▞▀       ▗▄▄▄▀▀▀▀▘▄▄▄▄▀▀▄▄▙▄▄▀▀▀▀▀▀▀▀▀▀▀│ 0.6
│     ▌         ▄▀▀      ▄▄▄▀▀▘▀▀▀                           │
│    ▗▘      ▄▄▀    ▗▄▗▄▀▀                                   │
│    ▐ ▗▀▀▀▛▀▘   ▄▞▗▞▀▘                                      │
│    ▐ ▐▗▄▀   ▗▞▀▄▞▘                                         │ 0.4
│  ▄▀▌▀▀▘   ▗▞▘▗▀                                            │
│▄▀  ▌ ▌  ▗▞▘ ▞▘                                             │
│   ▐ ▗▗▄▄▄▄▞▀                                               │
│   ▐▞▐▞▘                                                    │ 0.2
│▄▀▗▘ ▐                                                      │
│  ▌ ▄▘                                                      │
│▀▀▀▀▘                                                       │
└────────────────────────────────────────────────────────────┘
               10               20               30
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:25:54,967 - INFO - machamp.model.callback - Performance of 3.9990 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_36.pt
2025-07-10 11:25:56,763 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_35.pt
2025-07-10 11:25:56,890 - INFO - machamp.model.trainer - 
2025-07-10 11:25:56,890 - INFO - machamp.model.trainer - Epoch 37/60: training
2025-07-10 11:26:23,328 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:26:23,329 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.141459369817579e-05
2025-07-10 11:26:23,329 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.141459369817579e-05
2025-07-10 11:26:23,329 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 4.337545605306801e-06
2025-07-10 11:26:23,329 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:26:23,329 - INFO - machamp.model.trainer - Epoch 37: evaluating on dev
2025-07-10 11:26:24,908 - INFO - machamp.model.callback - epoch       : 37/60
2025-07-10 11:26:24,908 - INFO - machamp.model.callback - best_epoch  : 37
2025-07-10 11:26:24,909 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:26:24,909 - INFO - machamp.model.callback - cur_ram     : 2.1414
2025-07-10 11:26:24,909 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:26:24,909 - INFO - machamp.model.callback - time_total  : 0:18:15
2025-07-10 11:26:24,909 - INFO - machamp.model.callback - 
2025-07-10 11:26:24,909 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:26:24,909 - INFO - machamp.model.callback - Best (37)                                                  
2025-07-10 11:26:24,909 - INFO - machamp.model.callback - dependency_las     0.3423   2.0222       0.8337     0.6310 
2025-07-10 11:26:24,909 - INFO - machamp.model.callback - lemma_accuracy     0.5137   0.6279       0.8444     0.8393 
2025-07-10 11:26:24,909 - INFO - machamp.model.callback - morph_accuracy     0.9565   1.0741       0.7854     0.7437 
2025-07-10 11:26:24,909 - INFO - machamp.model.callback - upos_accuracy      0.0902   0.1422       0.9501     0.9114 
2025-07-10 11:26:24,909 - INFO - machamp.model.callback - xpos_accuracy      0.1424   0.1840       0.9235     0.8880 
2025-07-10 11:26:24,910 - INFO - machamp.model.callback - sum                2.0451   4.0504       4.3372     4.0134 
2025-07-10 11:26:24,910 - INFO - machamp.model.callback - Epoch 37                                                   
2025-07-10 11:26:24,910 - INFO - machamp.model.callback - dependency_las     0.3423   2.0222       0.8337     0.6310 
2025-07-10 11:26:24,910 - INFO - machamp.model.callback - lemma_accuracy     0.5137   0.6279       0.8444     0.8393 
2025-07-10 11:26:24,910 - INFO - machamp.model.callback - morph_accuracy     0.9565   1.0741       0.7854     0.7437 
2025-07-10 11:26:24,910 - INFO - machamp.model.callback - upos_accuracy      0.0902   0.1422       0.9501     0.9114 
2025-07-10 11:26:24,910 - INFO - machamp.model.callback - xpos_accuracy      0.1424   0.1840       0.9235     0.8880 
2025-07-10 11:26:24,910 - INFO - machamp.model.callback - sum                2.0451   4.0504       4.3372     4.0134 
2025-07-10 11:26:24,919 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                                 ▗▄▄▄▄▄▄▄▄▄▄▄▄▛▀▀▄▄▄▄▄▄▄▄▄▄▄│
│                           ▗▄▞▀▀▗▄▄▄▄▞▀▀▀▀▀▀▀▀▀▀▀▄▄▄▄▄▄▄▄▄▄▄│
│                        ▗▄▗▄▄▞▀▀▘▄▄▄▄▞▀▀▀▀▀▀▀▀▘             │ 0.8
│                    ▗▄▀▀▀▀▀▄▞▀▀▀                ▄▄▄▄▄▄▄▄▄▀▀▀│
│                 ▄▄▀▘▘▄▄▀▀          ▗▄▄▄▄▀▀▀▀▀▀▀            │
│     ▄▀▀▀▀▀▀▀▀▀▀▀▞▄▄▀▘       ▄▄▄▞▀▀▀▀▄▄▄▄▀▚▄▄▄▄▄▀▀▀▄▀▀▀▀▀▀▀▀│ 0.6
│     ▌        ▗▞▘▘     ▗▗▄▞▀▀▀▀▀                            │
│    ▐       ▄▟▘    ▄▄▄▞▀▘                                   │
│    ▐ ▐▀▀▗▀▀▘  ▗▄▞▄▞▀                                       │
│    ▞ ▐▗▄▘   ▗▞▗▄▀                                          │ 0.4
│ ▗▄▀▌▀▀▘   ▗▞▘▞▘                                            │
│▄▀ ▗▘ ▌  ▄▞▘▗▀                                              │
│   ▐ ▐▗▄▄▄▄▀▘                                               │
│   ▞▞▐▌                                                     │ 0.2
│▄▀▗▘ ▐                                                      │
│ ▗▌ ▞▘                                                      │
│▀▀▀▀                                                        │
└────────────────────────────────────────────────────────────┘
               10              20               30
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:26:24,920 - INFO - machamp.model.callback - Performance of 4.0134 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_37.pt
2025-07-10 11:26:26,733 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_36.pt
2025-07-10 11:26:26,859 - INFO - machamp.model.trainer - 
2025-07-10 11:26:26,860 - INFO - machamp.model.trainer - Epoch 38/60: training
2025-07-10 11:26:53,355 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:26:53,355 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.0945480928689884e-05
2025-07-10 11:26:53,356 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.0945480928689884e-05
2025-07-10 11:26:53,356 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 4.159282752902156e-06
2025-07-10 11:26:53,356 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:26:53,356 - INFO - machamp.model.trainer - Epoch 38: evaluating on dev
2025-07-10 11:26:54,916 - INFO - machamp.model.callback - epoch       : 38/60
2025-07-10 11:26:54,917 - INFO - machamp.model.callback - best_epoch  : 37
2025-07-10 11:26:54,917 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:26:54,917 - INFO - machamp.model.callback - cur_ram     : 2.1414
2025-07-10 11:26:54,917 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:26:54,917 - INFO - machamp.model.callback - time_total  : 0:18:45
2025-07-10 11:26:54,917 - INFO - machamp.model.callback - 
2025-07-10 11:26:54,917 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:26:54,917 - INFO - machamp.model.callback - Best (37)                                                  
2025-07-10 11:26:54,917 - INFO - machamp.model.callback - dependency_las     0.3423   2.0222       0.8337     0.6310 
2025-07-10 11:26:54,917 - INFO - machamp.model.callback - lemma_accuracy     0.5137   0.6279       0.8444     0.8393 
2025-07-10 11:26:54,917 - INFO - machamp.model.callback - morph_accuracy     0.9565   1.0741       0.7854     0.7437 
2025-07-10 11:26:54,917 - INFO - machamp.model.callback - upos_accuracy      0.0902   0.1422       0.9501     0.9114 
2025-07-10 11:26:54,917 - INFO - machamp.model.callback - xpos_accuracy      0.1424   0.1840       0.9235     0.8880 
2025-07-10 11:26:54,917 - INFO - machamp.model.callback - sum                2.0451   4.0504       4.3372     4.0134 
2025-07-10 11:26:54,917 - INFO - machamp.model.callback - Epoch 38                                                   
2025-07-10 11:26:54,917 - INFO - machamp.model.callback - dependency_las     0.3040   2.0454       0.8458     0.6219 
2025-07-10 11:26:54,917 - INFO - machamp.model.callback - lemma_accuracy     0.4994   0.6254       0.8486     0.8423 
2025-07-10 11:26:54,917 - INFO - machamp.model.callback - morph_accuracy     0.9268   1.0689       0.7893     0.7414 
2025-07-10 11:26:54,917 - INFO - machamp.model.callback - upos_accuracy      0.0843   0.1415       0.9525     0.9108 
2025-07-10 11:26:54,918 - INFO - machamp.model.callback - xpos_accuracy      0.1371   0.1826       0.9241     0.8863 
2025-07-10 11:26:54,918 - INFO - machamp.model.callback - sum                1.9516   4.0638       4.3604     4.0027 
2025-07-10 11:26:54,924 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                                ▗▄▄▄▄▄▄▄▄▄▄▄▄▀▀▗▄▄▄▄▄▄▄▄▄▄▄▄│
│                          ▗▄▞▀▀▗▄▄▄▄▞▀▀▀▀▀▀▀▀▀▀▘▄▄▄▄▄▄▄▄▄▄▄▄│
│                        ▄▞▄▄▞▀▀▘▄▄▄▄▞▀▀▀▀▀▀▀▀               │ 0.8
│                   ▗▄▛▀▀▀▀▄▟▀▀▀               ▗▄▄▄▄▄▄▄▄▞▀▀▀▀│
│                ▗▄▞▀▀▄▄▞▀▘         ▗▄▄▄▄▀▀▀▀▀▀▘             │
│     ▄▀▀▀▀▀▀▀▀▀▀▘▘▄▞▀       ▄▄▄▞▀▀▀▀▄▄▄▄▀▄▄▄▄▄▞▀▀▚▞▀▀▀▀▀▀▀▀▚│ 0.6
│    ▗▘        ▄▀▀▘     ▄▄▞▀▀▀▀▀▘                            │
│    ▐      ▗▄▛    ▗▄▗▄▀▀                                    │
│    ▐ ▐▀▀▐▀▀   ▄▄▗▄▀▘                                       │
│    ▌ ▌▄▞▘   ▄▀▗▞▘                                          │ 0.4
│ ▗▄▀▌▀▀    ▄▀▗▀▘                                            │
│▄▀ ▐ ▗▘  ▄▀ ▄▘                                              │
│   ▐ ▐▗▄▄▄▄▀                                                │
│  ▗▞▀▗▘                                                     │ 0.2
│▄▀▗▘ ▞                                                      │
│ ▗▌▗▞▘                                                      │
│▀▀▀▀                                                        │
└────────────────────────────────────────────────────────────┘
              10              20              30
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:26:54,924 - INFO - machamp.model.trainer - 
2025-07-10 11:26:54,924 - INFO - machamp.model.trainer - Epoch 39/60: training
2025-07-10 11:27:21,450 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:27:21,451 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.0476368159203982e-05
2025-07-10 11:27:21,451 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.0476368159203982e-05
2025-07-10 11:27:21,451 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 3.9810199004975135e-06
2025-07-10 11:27:21,451 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:27:21,451 - INFO - machamp.model.trainer - Epoch 39: evaluating on dev
2025-07-10 11:27:23,001 - INFO - machamp.model.callback - epoch       : 39/60
2025-07-10 11:27:23,001 - INFO - machamp.model.callback - best_epoch  : 39
2025-07-10 11:27:23,001 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:27:23,001 - INFO - machamp.model.callback - cur_ram     : 2.1414
2025-07-10 11:27:23,001 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:27:23,001 - INFO - machamp.model.callback - time_total  : 0:19:13
2025-07-10 11:27:23,001 - INFO - machamp.model.callback - 
2025-07-10 11:27:23,001 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:27:23,001 - INFO - machamp.model.callback - Best (39)                                                  
2025-07-10 11:27:23,001 - INFO - machamp.model.callback - dependency_las     0.3055   2.0719       0.8476     0.6280 
2025-07-10 11:27:23,001 - INFO - machamp.model.callback - lemma_accuracy     0.4925   0.6170       0.8509     0.8427 
2025-07-10 11:27:23,001 - INFO - machamp.model.callback - morph_accuracy     0.9084   1.0503       0.7945     0.7464 
2025-07-10 11:27:23,001 - INFO - machamp.model.callback - upos_accuracy      0.0815   0.1401       0.9548     0.9114 
2025-07-10 11:27:23,001 - INFO - machamp.model.callback - xpos_accuracy      0.1322   0.1812       0.9271     0.8883 
2025-07-10 11:27:23,001 - INFO - machamp.model.callback - sum                1.9201   4.0605       4.3748     4.0168 
2025-07-10 11:27:23,001 - INFO - machamp.model.callback - Epoch 39                                                   
2025-07-10 11:27:23,001 - INFO - machamp.model.callback - dependency_las     0.3055   2.0719       0.8476     0.6280 
2025-07-10 11:27:23,001 - INFO - machamp.model.callback - lemma_accuracy     0.4925   0.6170       0.8509     0.8427 
2025-07-10 11:27:23,001 - INFO - machamp.model.callback - morph_accuracy     0.9084   1.0503       0.7945     0.7464 
2025-07-10 11:27:23,002 - INFO - machamp.model.callback - upos_accuracy      0.0815   0.1401       0.9548     0.9114 
2025-07-10 11:27:23,002 - INFO - machamp.model.callback - xpos_accuracy      0.1322   0.1812       0.9271     0.8883 
2025-07-10 11:27:23,002 - INFO - machamp.model.callback - sum                1.9201   4.0605       4.3748     4.0168 
2025-07-10 11:27:23,008 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                                ▄▄▄▄▄▄▄▄▄▄▄▄▀▀▄▄▄▄▄▄▄▄▄▄▄▄▄▄│
│                          ▄▄▀▀▀▄▄▄▄▄▀▀▀▀▀▀▀▀▀▀▄▄▄▄▄▄▄▄▄▄▄▄▄▄│
│                       ▗▞▗▄▄▀▀▀▄▄▄▄▞▀▀▀▀▀▀▀▀                │ 0.8
│                   ▄▟▀▀▀▀▘▄▀▀▀▘              ▗▄▄▄▄▄▄▄▄▀▀▀▀▀▀│
│                ▄▄▀▘▘▄▞▀▘          ▄▄▄▄▀▀▀▀▀▀▘              │
│    ▗▄▀▀▀▀▀▀▀▀▀▀▞▄▄▀▘      ▗▄▄▞▀▀▀▀▄▄▄▄▀▄▄▄▄▄▞▀▀▚▞▀▀▀▀▀▀▀▚▄▀│ 0.6
│    ▐        ▗▄▘▀     ▗▗▄▀▀▘▀▀▘                             │
│    ▐      ▗▄▀    ▄▄▄▞▀▘                                    │
│    ▞ ▞▀▀▞▀▘  ▗▄▞▄▞▀                                        │
│    ▌ ▌▄▞   ▗▞▘▄▞                                           │ 0.4
│ ▗▄▀▌▀▀   ▗▞▘▄▀                                             │
│▄▀ ▐ ▐  ▗▄▘ ▞                                               │
│   ▐ ▐▄▄▄▄▞▀                                                │
│  ▗▌▀▗▘                                                     │ 0.2
│▄▀▐ ▗▞                                                      │
│ ▗▌▗▞                                                       │
│▀▀▀▘                                                        │
└────────────────────────────────────────────────────────────┘
              10              20             30
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:27:23,008 - INFO - machamp.model.callback - Performance of 4.0168 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_39.pt
2025-07-10 11:27:24,954 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_37.pt
2025-07-10 11:27:25,095 - INFO - machamp.model.trainer - 
2025-07-10 11:27:25,095 - INFO - machamp.model.trainer - Epoch 40/60: training
2025-07-10 11:27:51,698 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:27:51,699 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.0007255389718079e-05
2025-07-10 11:27:51,699 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.0007255389718079e-05
2025-07-10 11:27:51,699 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 3.80275704809287e-06
2025-07-10 11:27:51,699 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:27:51,699 - INFO - machamp.model.trainer - Epoch 40: evaluating on dev
2025-07-10 11:27:53,269 - INFO - machamp.model.callback - epoch       : 40/60
2025-07-10 11:27:53,269 - INFO - machamp.model.callback - best_epoch  : 39
2025-07-10 11:27:53,269 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:27:53,269 - INFO - machamp.model.callback - cur_ram     : 2.1279
2025-07-10 11:27:53,269 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:27:53,269 - INFO - machamp.model.callback - time_total  : 0:19:43
2025-07-10 11:27:53,270 - INFO - machamp.model.callback - 
2025-07-10 11:27:53,270 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:27:53,270 - INFO - machamp.model.callback - Best (39)                                                  
2025-07-10 11:27:53,270 - INFO - machamp.model.callback - dependency_las     0.3055   2.0719       0.8476     0.6280 
2025-07-10 11:27:53,270 - INFO - machamp.model.callback - lemma_accuracy     0.4925   0.6170       0.8509     0.8427 
2025-07-10 11:27:53,270 - INFO - machamp.model.callback - morph_accuracy     0.9084   1.0503       0.7945     0.7464 
2025-07-10 11:27:53,270 - INFO - machamp.model.callback - upos_accuracy      0.0815   0.1401       0.9548     0.9114 
2025-07-10 11:27:53,270 - INFO - machamp.model.callback - xpos_accuracy      0.1322   0.1812       0.9271     0.8883 
2025-07-10 11:27:53,270 - INFO - machamp.model.callback - sum                1.9201   4.0605       4.3748     4.0168 
2025-07-10 11:27:53,270 - INFO - machamp.model.callback - Epoch 40                                                   
2025-07-10 11:27:53,270 - INFO - machamp.model.callback - dependency_las     0.2801   2.0517       0.8465     0.6246 
2025-07-10 11:27:53,270 - INFO - machamp.model.callback - lemma_accuracy     0.4828   0.6113       0.8546     0.8437 
2025-07-10 11:27:53,270 - INFO - machamp.model.callback - morph_accuracy     0.8851   1.0402       0.7997     0.7471 
2025-07-10 11:27:53,271 - INFO - machamp.model.callback - upos_accuracy      0.0797   0.1419       0.9544     0.9084 
2025-07-10 11:27:53,271 - INFO - machamp.model.callback - xpos_accuracy      0.1290   0.1808       0.9293     0.8886 
2025-07-10 11:27:53,271 - INFO - machamp.model.callback - sum                1.8566   4.0260       4.3845     4.0124 
2025-07-10 11:27:53,277 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                               ▄▄▄▄▄▄▄▄▄▄▄▟▀▀▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄│
│                         ▗▄▀▀▀▄▄▄▄▄▀▀▀▀▀▀▀▀▀▀▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄│
│                      ▗▄▗▄▄▀▀▀▄▄▄▄▞▀▀▀▀▀▀▀▘                 │ 0.8
│                  ▗▄▛▀▀▀▀▄▞▀▀▘              ▄▄▄▄▄▄▄▄▞▀▀▀▀▀▀▀│
│               ▗▄▞▀▀▄▄▀▀          ▄▄▄▄▀▀▀▀▀▀▘               │
│    ▗▄▀▀▀▀▀▀▀▀▀▀▀▄▞▀      ▗▄▄▄▀▀▀▀▄▄▄▄▀▄▄▄▄▄▀▀▀▄▀▀▀▀▀▀▀▀▄▞▀▀│ 0.6
│    ▐        ▗▞▘▘     ▗▄▞▀▘▀▀▘                              │
│    ▐      ▄▞▘   ▗▄▗▄▀▘                                     │
│    ▌ ▛▀▀▛▀▘  ▗▄▗▄▀▘                                        │
│    ▌ ▌▄▞   ▄▀▗▄▘                                           │ 0.4
│ ▗▞▗▘▀▀   ▗▞▘▞▘                                             │
│▄▀ ▐ ▐  ▗▞▘▗▀                                               │
│   ▐ ▌▄▄▄▄▞▘                                                │
│  ▗▌▀▐▘                                                     │ 0.2
│▄▀▐ ▗▌                                                      │
│ ▗▘▗▞                                                       │
│▀▘▀▘                                                        │
└────────────────────────────────────────────────────────────┘
             10              20             30             40
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:27:53,278 - INFO - machamp.model.trainer - 
2025-07-10 11:27:53,278 - INFO - machamp.model.trainer - Epoch 41/60: training
2025-07-10 11:28:19,710 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:28:19,710 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 9.538142620232174e-06
2025-07-10 11:28:19,710 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 9.538142620232174e-06
2025-07-10 11:28:19,710 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 3.624494195688226e-06
2025-07-10 11:28:19,710 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:28:19,711 - INFO - machamp.model.trainer - Epoch 41: evaluating on dev
2025-07-10 11:28:21,274 - INFO - machamp.model.callback - epoch       : 41/60
2025-07-10 11:28:21,274 - INFO - machamp.model.callback - best_epoch  : 41
2025-07-10 11:28:21,274 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:28:21,275 - INFO - machamp.model.callback - cur_ram     : 2.1279
2025-07-10 11:28:21,275 - INFO - machamp.model.callback - time_epoch  : 0:00:27
2025-07-10 11:28:21,275 - INFO - machamp.model.callback - time_total  : 0:20:11
2025-07-10 11:28:21,275 - INFO - machamp.model.callback - 
2025-07-10 11:28:21,275 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:28:21,275 - INFO - machamp.model.callback - Best (41)                                                  
2025-07-10 11:28:21,275 - INFO - machamp.model.callback - dependency_las     0.2598   2.1019       0.8568     0.6283 
2025-07-10 11:28:21,275 - INFO - machamp.model.callback - lemma_accuracy     0.4753   0.6048       0.8590     0.8457 
2025-07-10 11:28:21,275 - INFO - machamp.model.callback - morph_accuracy     0.8659   1.0282       0.8031     0.7484 
2025-07-10 11:28:21,275 - INFO - machamp.model.callback - upos_accuracy      0.0763   0.1398       0.9577     0.9145 
2025-07-10 11:28:21,275 - INFO - machamp.model.callback - xpos_accuracy      0.1240   0.1790       0.9335     0.8900 
2025-07-10 11:28:21,275 - INFO - machamp.model.callback - sum                1.8014   4.0536       4.4100     4.0268 
2025-07-10 11:28:21,275 - INFO - machamp.model.callback - Epoch 41                                                   
2025-07-10 11:28:21,275 - INFO - machamp.model.callback - dependency_las     0.2598   2.1019       0.8568     0.6283 
2025-07-10 11:28:21,275 - INFO - machamp.model.callback - lemma_accuracy     0.4753   0.6048       0.8590     0.8457 
2025-07-10 11:28:21,275 - INFO - machamp.model.callback - morph_accuracy     0.8659   1.0282       0.8031     0.7484 
2025-07-10 11:28:21,275 - INFO - machamp.model.callback - upos_accuracy      0.0763   0.1398       0.9577     0.9145 
2025-07-10 11:28:21,275 - INFO - machamp.model.callback - xpos_accuracy      0.1240   0.1790       0.9335     0.8900 
2025-07-10 11:28:21,275 - INFO - machamp.model.callback - sum                1.8014   4.0536       4.4100     4.0268 
2025-07-10 11:28:21,284 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                               ▄▄▄▄▄▄▄▄▄▄▄▄▄▀▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄│
│                         ▄▄▀▀▗▄▄▄▄▞▀▀▀▀▀▀▀▀▀▀▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄│
│                      ▗▞▗▄▞▀▀▘▄▄▄▄▀▀▀▀▀▀▀▀▘                 │ 0.8
│                  ▗▄▀▀▀▀▘▄▀▀▀               ▄▄▄▄▄▄▄▄▀▀▀▀▀▀▀▀│
│               ▗▄▞▘▀▄▞▀▘         ▗▄▄▄▞▀▀▀▀▀▀                │
│     ▞▀▀▀▀▀▀▀▀▀▘▘▄▀▀      ▄▄▄▞▀▀▀▘▄▄▄▀▄▄▄▄▄▞▀▚▄▞▀▀▀▀▀▀▚▄▞▚▄▀│ 0.6
│    ▐        ▗▀▀     ▗▄▄▀▀▀▀▀▘                              │
│    ▞      ▄▞▘   ▗▄▄▞▀                                      │
│    ▌ ▛▀▗▀▀   ▄▞▄▞▀                                         │
│    ▌▗▘▄▘   ▄▀▗▞                                            │ 0.4
│ ▗▞▐▀▀▀   ▄▀▗▀▘                                             │
│▄▀ ▐ ▞  ▗▞ ▄▘                                               │
│   ▞ ▌▄▄▄▄▀                                                 │
│  ▗▌▀▐▘                                                     │ 0.2
│▄▀▐ ▐▌                                                      │
│ ▗▘▗▀                                                       │
│▀▘▀▘                                                        │
└────────────────────────────────────────────────────────────┘
             10             20             30             40
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:28:21,285 - INFO - machamp.model.callback - Performance of 4.0268 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_41.pt
2025-07-10 11:28:23,067 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_39.pt
2025-07-10 11:28:23,196 - INFO - machamp.model.trainer - 
2025-07-10 11:28:23,197 - INFO - machamp.model.trainer - Epoch 42/60: training
2025-07-10 11:28:49,854 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:28:49,854 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 9.069029850746269e-06
2025-07-10 11:28:49,854 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 9.069029850746269e-06
2025-07-10 11:28:49,854 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 3.446231343283582e-06
2025-07-10 11:28:49,854 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:28:49,854 - INFO - machamp.model.trainer - Epoch 42: evaluating on dev
2025-07-10 11:28:51,438 - INFO - machamp.model.callback - epoch       : 42/60
2025-07-10 11:28:51,438 - INFO - machamp.model.callback - best_epoch  : 42
2025-07-10 11:28:51,438 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:28:51,438 - INFO - machamp.model.callback - cur_ram     : 2.1409
2025-07-10 11:28:51,438 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:28:51,438 - INFO - machamp.model.callback - time_total  : 0:20:42
2025-07-10 11:28:51,438 - INFO - machamp.model.callback - 
2025-07-10 11:28:51,439 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:28:51,439 - INFO - machamp.model.callback - Best (42)                                                  
2025-07-10 11:28:51,439 - INFO - machamp.model.callback - dependency_las     0.2497   2.1039       0.8617     0.6290 
2025-07-10 11:28:51,439 - INFO - machamp.model.callback - lemma_accuracy     0.4636   0.6000       0.8584     0.8494 
2025-07-10 11:28:51,439 - INFO - machamp.model.callback - morph_accuracy     0.8465   1.0181       0.8067     0.7518 
2025-07-10 11:28:51,439 - INFO - machamp.model.callback - upos_accuracy      0.0729   0.1388       0.9596     0.9114 
2025-07-10 11:28:51,439 - INFO - machamp.model.callback - xpos_accuracy      0.1195   0.1779       0.9360     0.8896 
2025-07-10 11:28:51,439 - INFO - machamp.model.callback - sum                1.7523   4.0388       4.4223     4.0312 
2025-07-10 11:28:51,439 - INFO - machamp.model.callback - Epoch 42                                                   
2025-07-10 11:28:51,439 - INFO - machamp.model.callback - dependency_las     0.2497   2.1039       0.8617     0.6290 
2025-07-10 11:28:51,439 - INFO - machamp.model.callback - lemma_accuracy     0.4636   0.6000       0.8584     0.8494 
2025-07-10 11:28:51,439 - INFO - machamp.model.callback - morph_accuracy     0.8465   1.0181       0.8067     0.7518 
2025-07-10 11:28:51,439 - INFO - machamp.model.callback - upos_accuracy      0.0729   0.1388       0.9596     0.9114 
2025-07-10 11:28:51,439 - INFO - machamp.model.callback - xpos_accuracy      0.1195   0.1779       0.9360     0.8896 
2025-07-10 11:28:51,440 - INFO - machamp.model.callback - sum                1.7523   4.0388       4.4223     4.0312 
2025-07-10 11:28:51,446 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                              ▄▄▄▄▄▄▄▄▄▄▄▄▞▀▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄│
│                        ▗▄▀▀▀▄▄▄▄▄▀▀▀▀▀▀▀▀▀▀▄▄▄▄▄▄▄▄▄▄▄▄▄▄▞▀│
│                      ▄▗▄▄▀▀▀▄▄▄▄▀▀▀▀▀▀▀▀▘                  │ 0.8
│                  ▄▞▀▀▀▘▄▀▀▀▘              ▄▄▄▄▄▄▄▞▀▀▀▀▀▀▀▀▀│
│               ▄▄▀▀▘▄▀▀         ▗▄▄▄▟▀▀▀▀▀▀                 │
│    ▗▞▀▀▀▀▀▀▀▀▀▞▄▞▀▘     ▗▄▄▞▀▀▀▘▄▄▄▞▄▄▄▄▄▞▀▚▄▞▀▀▀▀▀▀▄▄▀▄▞▀▀│ 0.6
│    ▐        ▄▘▘     ▗▄▀▀▘▀▀▘                               │
│    ▌     ▗▄▀    ▄▗▄▀▘                                      │
│    ▌ ▛▀▐▀▘  ▗▄▞▄▀▘                                         │
│   ▗▘▐▗▄▘  ▗▞▘▄▀                                            │ 0.4
│ ▗▞▐▀▀▘   ▄▀▗▀                                              │
│▄▘ ▐ ▌  ▄▀ ▞▘                                               │
│   ▌ ▗▄▄▄▄▀                                                 │
│  ▗▌▗▞▘                                                     │ 0.2
│▄▀▞ ▗▘                                                      │
│ ▗▘▗▀                                                       │
│▀▘▀▘                                                        │
└────────────────────────────────────────────────────────────┘
             10            20             30             40
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:28:51,446 - INFO - machamp.model.callback - Performance of 4.0312 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_42.pt
2025-07-10 11:28:53,209 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_41.pt
2025-07-10 11:28:53,341 - INFO - machamp.model.trainer - 
2025-07-10 11:28:53,341 - INFO - machamp.model.trainer - Epoch 43/60: training
2025-07-10 11:29:19,971 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:29:19,971 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 8.599917081260365e-06
2025-07-10 11:29:19,971 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 8.599917081260365e-06
2025-07-10 11:29:19,971 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 3.2679684908789387e-06
2025-07-10 11:29:19,971 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:29:19,971 - INFO - machamp.model.trainer - Epoch 43: evaluating on dev
2025-07-10 11:29:21,516 - INFO - machamp.model.callback - epoch       : 43/60
2025-07-10 11:29:21,516 - INFO - machamp.model.callback - best_epoch  : 43
2025-07-10 11:29:21,516 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:29:21,516 - INFO - machamp.model.callback - cur_ram     : 2.1279
2025-07-10 11:29:21,516 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:29:21,516 - INFO - machamp.model.callback - time_total  : 0:21:12
2025-07-10 11:29:21,517 - INFO - machamp.model.callback - 
2025-07-10 11:29:21,517 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:29:21,517 - INFO - machamp.model.callback - Best (43)                                                  
2025-07-10 11:29:21,517 - INFO - machamp.model.callback - dependency_las     0.2421   2.1299       0.8611     0.6327 
2025-07-10 11:29:21,517 - INFO - machamp.model.callback - lemma_accuracy     0.4592   0.5958       0.8602     0.8504 
2025-07-10 11:29:21,517 - INFO - machamp.model.callback - morph_accuracy     0.8362   1.0097       0.8094     0.7544 
2025-07-10 11:29:21,517 - INFO - machamp.model.callback - upos_accuracy      0.0700   0.1406       0.9613     0.9124 
2025-07-10 11:29:21,517 - INFO - machamp.model.callback - xpos_accuracy      0.1171   0.1788       0.9397     0.8896 
2025-07-10 11:29:21,517 - INFO - machamp.model.callback - sum                1.7247   4.0548       4.4317     4.0396 
2025-07-10 11:29:21,517 - INFO - machamp.model.callback - Epoch 43                                                   
2025-07-10 11:29:21,517 - INFO - machamp.model.callback - dependency_las     0.2421   2.1299       0.8611     0.6327 
2025-07-10 11:29:21,517 - INFO - machamp.model.callback - lemma_accuracy     0.4592   0.5958       0.8602     0.8504 
2025-07-10 11:29:21,517 - INFO - machamp.model.callback - morph_accuracy     0.8362   1.0097       0.8094     0.7544 
2025-07-10 11:29:21,518 - INFO - machamp.model.callback - upos_accuracy      0.0700   0.1406       0.9613     0.9124 
2025-07-10 11:29:21,518 - INFO - machamp.model.callback - xpos_accuracy      0.1171   0.1788       0.9397     0.8896 
2025-07-10 11:29:21,518 - INFO - machamp.model.callback - sum                1.7247   4.0548       4.4317     4.0396 
2025-07-10 11:29:21,525 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                             ▗▄▄▄▄▄▄▄▄▄▄▄▞▀▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄│
│                       ▗▄▞▀▀▄▄▄▄▄▀▀▀▀▀▀▀▀▀▀▄▄▄▄▄▄▄▄▄▄▄▄▄▄▀▀▀│
│                     ▗▞▄▄▞▀▀▄▄▄▄▞▀▀▀▀▀▀▀▘                   │ 0.8
│                 ▗▄▀▀▀▀▄▞▀▀▀              ▄▄▄▄▄▄▄▞▀▀▀▀▀▀▀▀▀▀│
│              ▗▄▞▘▀▄▞▀▘         ▄▄▄▄▀▀▀▀▀▀                  │
│    ▗▞▀▀▀▀▀▀▀▀▘▘▄▀▀      ▄▄▄▀▀▀▀▄▄▄▞▄▄▄▄▄▞▀▚▄▞▀▀▀▀▀▀▄▞▚▄▀▀▀▀│ 0.6
│    ▞       ▗▞▘▘    ▗▄▞▀▀▀▀▀                                │
│    ▌     ▗▟▘   ▗▄▄▞▀                                       │
│    ▌▐▀▀▐▀▘  ▗▄▗▞▀                                          │
│   ▐ ▐▗▄▘  ▗▞▗▄▘                                            │ 0.4
│ ▗▞▐▀▀▘  ▗▞▘▞▘                                              │
│▄▘ ▞ ▌  ▄▘▗▞                                                │
│   ▌ ▗▄▄▄▞▘                                                 │
│  ▗▘▐▞                                                      │ 0.2
│▄▀▞ ▗▘                                                      │
│ ▗▘▗▀                                                       │
│▀▘▀▘                                                        │
└────────────────────────────────────────────────────────────┘
            10             20            30            40
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:29:21,525 - INFO - machamp.model.callback - Performance of 4.0396 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_43.pt
2025-07-10 11:29:23,327 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_42.pt
2025-07-10 11:29:23,458 - INFO - machamp.model.trainer - 
2025-07-10 11:29:23,458 - INFO - machamp.model.trainer - Epoch 44/60: training
2025-07-10 11:29:49,949 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:29:49,949 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 8.130804311774462e-06
2025-07-10 11:29:49,949 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 8.130804311774462e-06
2025-07-10 11:29:49,949 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 3.0897056384742954e-06
2025-07-10 11:29:49,950 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:29:49,950 - INFO - machamp.model.trainer - Epoch 44: evaluating on dev
2025-07-10 11:29:51,524 - INFO - machamp.model.callback - epoch       : 44/60
2025-07-10 11:29:51,524 - INFO - machamp.model.callback - best_epoch  : 44
2025-07-10 11:29:51,524 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:29:51,524 - INFO - machamp.model.callback - cur_ram     : 2.1409
2025-07-10 11:29:51,524 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:29:51,524 - INFO - machamp.model.callback - time_total  : 0:21:42
2025-07-10 11:29:51,524 - INFO - machamp.model.callback - 
2025-07-10 11:29:51,524 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:29:51,524 - INFO - machamp.model.callback - Best (44)                                                  
2025-07-10 11:29:51,524 - INFO - machamp.model.callback - dependency_las     0.2427   2.1486       0.8700     0.6317 
2025-07-10 11:29:51,524 - INFO - machamp.model.callback - lemma_accuracy     0.4537   0.5946       0.8624     0.8527 
2025-07-10 11:29:51,525 - INFO - machamp.model.callback - morph_accuracy     0.8230   1.0022       0.8131     0.7565 
2025-07-10 11:29:51,525 - INFO - machamp.model.callback - upos_accuracy      0.0702   0.1411       0.9620     0.9101 
2025-07-10 11:29:51,525 - INFO - machamp.model.callback - xpos_accuracy      0.1180   0.1782       0.9352     0.8920 
2025-07-10 11:29:51,525 - INFO - machamp.model.callback - sum                1.7076   4.0646       4.4428     4.0429 
2025-07-10 11:29:51,525 - INFO - machamp.model.callback - Epoch 44                                                   
2025-07-10 11:29:51,525 - INFO - machamp.model.callback - dependency_las     0.2427   2.1486       0.8700     0.6317 
2025-07-10 11:29:51,525 - INFO - machamp.model.callback - lemma_accuracy     0.4537   0.5946       0.8624     0.8527 
2025-07-10 11:29:51,525 - INFO - machamp.model.callback - morph_accuracy     0.8230   1.0022       0.8131     0.7565 
2025-07-10 11:29:51,525 - INFO - machamp.model.callback - upos_accuracy      0.0702   0.1411       0.9620     0.9101 
2025-07-10 11:29:51,525 - INFO - machamp.model.callback - xpos_accuracy      0.1180   0.1782       0.9352     0.8920 
2025-07-10 11:29:51,525 - INFO - machamp.model.callback - sum                1.7076   4.0646       4.4428     4.0429 
2025-07-10 11:29:51,531 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                            ▗▄▄▄▄▄▄▄▄▄▄▄▞▀▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄│
│                       ▄▄▀▀▗▄▄▄▄▞▀▀▀▀▀▀▀▀▀▄▄▄▄▄▄▄▄▄▄▄▄▄▄▀▀▀▀│
│                     ▄▗▄▄▀▀▘▄▄▄▞▀▀▀▀▀▀▀▘                    │ 0.8
│                 ▄▞▀▀▀▘▄▀▀▀              ▄▄▄▄▄▄▄▀▀▀▀▀▀▀▀▀▀▀▀│
│              ▗▄▀▀▘▄▀▀         ▄▄▄▄▀▀▀▀▀▀                   │
│    ▗▞▀▀▀▀▀▀▀▀▘▘▟▀▘     ▗▄▄▞▀▀▀▄▄▄▞▄▄▄▄▄▞▀▚▄▞▀▀▀▀▀▚▄▞▚▄▀▀▀▀▀│ 0.6
│    ▌       ▗▀▀     ▗▄▀▀▘▀▀▘                                │
│    ▌     ▄▞▘   ▄▄▄▀▘                                       │
│   ▗▘▐▀▀▞▀▘  ▄▞▄▞▀                                          │
│   ▐ ▐▗▞   ▄▀▗▞                                             │ 0.4
│ ▗▞▐▀▀▘  ▗▞▗▞▘                                              │
│▄▘ ▌ ▌ ▗▟▘▗▘                                                │
│   ▌▗▗▄▄▄▞▘                                                 │
│  ▗▘▐▌                                                      │ 0.2
│▄▀▞ ▐                                                       │
│ ▗▘▄▘                                                       │
│▀▘▀▘                                                        │
└────────────────────────────────────────────────────────────┘
            10            20            30            40
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:29:51,531 - INFO - machamp.model.callback - Performance of 4.0429 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_44.pt
2025-07-10 11:29:53,301 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_43.pt
2025-07-10 11:29:53,431 - INFO - machamp.model.trainer - 
2025-07-10 11:29:53,431 - INFO - machamp.model.trainer - Epoch 45/60: training
2025-07-10 11:30:19,943 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:30:19,943 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 7.661691542288559e-06
2025-07-10 11:30:19,943 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 7.661691542288559e-06
2025-07-10 11:30:19,943 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 2.9114427860696525e-06
2025-07-10 11:30:19,943 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:30:19,944 - INFO - machamp.model.trainer - Epoch 45: evaluating on dev
2025-07-10 11:30:21,504 - INFO - machamp.model.callback - epoch       : 45/60
2025-07-10 11:30:21,504 - INFO - machamp.model.callback - best_epoch  : 45
2025-07-10 11:30:21,504 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:30:21,504 - INFO - machamp.model.callback - cur_ram     : 2.1409
2025-07-10 11:30:21,504 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:30:21,504 - INFO - machamp.model.callback - time_total  : 0:22:12
2025-07-10 11:30:21,504 - INFO - machamp.model.callback - 
2025-07-10 11:30:21,504 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:30:21,504 - INFO - machamp.model.callback - Best (45)                                                  
2025-07-10 11:30:21,505 - INFO - machamp.model.callback - dependency_las     0.2317   2.1903       0.8716     0.6276 
2025-07-10 11:30:21,505 - INFO - machamp.model.callback - lemma_accuracy     0.4447   0.5887       0.8673     0.8544 
2025-07-10 11:30:21,505 - INFO - machamp.model.callback - morph_accuracy     0.8084   0.9911       0.8153     0.7615 
2025-07-10 11:30:21,505 - INFO - machamp.model.callback - upos_accuracy      0.0673   0.1389       0.9613     0.9155 
2025-07-10 11:30:21,505 - INFO - machamp.model.callback - xpos_accuracy      0.1107   0.1774       0.9397     0.8940 
2025-07-10 11:30:21,505 - INFO - machamp.model.callback - sum                1.6628   4.0863       4.4553     4.0530 
2025-07-10 11:30:21,505 - INFO - machamp.model.callback - Epoch 45                                                   
2025-07-10 11:30:21,505 - INFO - machamp.model.callback - dependency_las     0.2317   2.1903       0.8716     0.6276 
2025-07-10 11:30:21,505 - INFO - machamp.model.callback - lemma_accuracy     0.4447   0.5887       0.8673     0.8544 
2025-07-10 11:30:21,505 - INFO - machamp.model.callback - morph_accuracy     0.8084   0.9911       0.8153     0.7615 
2025-07-10 11:30:21,505 - INFO - machamp.model.callback - upos_accuracy      0.0673   0.1389       0.9613     0.9155 
2025-07-10 11:30:21,505 - INFO - machamp.model.callback - xpos_accuracy      0.1107   0.1774       0.9397     0.8940 
2025-07-10 11:30:21,505 - INFO - machamp.model.callback - sum                1.6628   4.0863       4.4553     4.0530 
2025-07-10 11:30:21,514 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                            ▗▄▄▄▄▄▄▄▄▄▄▄▞▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄│
│                      ▗▄▞▀▀▄▄▄▄▄▀▀▀▀▀▀▀▀▀▄▄▄▄▄▄▄▄▄▄▄▄▄▄▀▀▀▀▀│
│                    ▗▞▄▄▞▀▀▄▄▄▄▀▀▀▀▀▀▀▀                     │ 0.8
│                 ▄▀▀▀▀▄▞▀▀▘             ▄▄▄▄▄▄▄▞▀▀▀▀▀▀▀▀▀▀▀▀│
│              ▄▄▀▀▄▄▀▘        ▗▄▄▄▞▀▀▀▀▀                    │
│    ▗▀▀▀▀▀▀▀▀▀▞▄▟▀      ▄▄▄▀▀▀▘▄▄▄▚▄▄▄▄▄▀▄▄▄▀▀▀▟▀▚▄▞▄▄▀▀▀▀▀▀│ 0.6
│    ▌       ▄▘▘    ▗▗▞▀▀▀▀▘                                 │
│    ▌     ▄▛    ▄▗▞▀▘                                       │
│   ▐ ▐▀▀▛▀   ▄▞▄▀▘                                          │
│   ▐ ▞▗▞   ▄▀▄▀                                             │ 0.4
│ ▗▞▐▀▀▘  ▗▀▗▀                                               │
│▄▀ ▌ ▌ ▗▞▘▄▘                                                │
│   ▌▐▗▄▄▄▞                                                  │
│  ▗▘▐▌                                                      │ 0.2
│▄▀▞ ▐                                                       │
│ ▗▘▄▘                                                       │
│▀▘▀▘                                                        │
└────────────────────────────────────────────────────────────┘
            10           20            30            40
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:30:21,515 - INFO - machamp.model.callback - Performance of 4.0530 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_45.pt
2025-07-10 11:30:23,299 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_44.pt
2025-07-10 11:30:23,428 - INFO - machamp.model.trainer - 
2025-07-10 11:30:23,428 - INFO - machamp.model.trainer - Epoch 46/60: training
2025-07-10 11:30:50,030 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:30:50,030 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 7.192578772802654e-06
2025-07-10 11:30:50,030 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 7.192578772802654e-06
2025-07-10 11:30:50,030 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 2.733179933665009e-06
2025-07-10 11:30:50,030 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:30:50,030 - INFO - machamp.model.trainer - Epoch 46: evaluating on dev
2025-07-10 11:30:51,623 - INFO - machamp.model.callback - epoch       : 46/60
2025-07-10 11:30:51,623 - INFO - machamp.model.callback - best_epoch  : 45
2025-07-10 11:30:51,623 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:30:51,623 - INFO - machamp.model.callback - cur_ram     : 2.1410
2025-07-10 11:30:51,623 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:30:51,623 - INFO - machamp.model.callback - time_total  : 0:22:42
2025-07-10 11:30:51,623 - INFO - machamp.model.callback - 
2025-07-10 11:30:51,623 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:30:51,623 - INFO - machamp.model.callback - Best (45)                                                  
2025-07-10 11:30:51,623 - INFO - machamp.model.callback - dependency_las     0.2317   2.1903       0.8716     0.6276 
2025-07-10 11:30:51,623 - INFO - machamp.model.callback - lemma_accuracy     0.4447   0.5887       0.8673     0.8544 
2025-07-10 11:30:51,623 - INFO - machamp.model.callback - morph_accuracy     0.8084   0.9911       0.8153     0.7615 
2025-07-10 11:30:51,623 - INFO - machamp.model.callback - upos_accuracy      0.0673   0.1389       0.9613     0.9155 
2025-07-10 11:30:51,623 - INFO - machamp.model.callback - xpos_accuracy      0.1107   0.1774       0.9397     0.8940 
2025-07-10 11:30:51,623 - INFO - machamp.model.callback - sum                1.6628   4.0863       4.4553     4.0530 
2025-07-10 11:30:51,623 - INFO - machamp.model.callback - Epoch 46                                                   
2025-07-10 11:30:51,623 - INFO - machamp.model.callback - dependency_las     0.2192   2.1890       0.8754     0.6286 
2025-07-10 11:30:51,623 - INFO - machamp.model.callback - lemma_accuracy     0.4346   0.5872       0.8664     0.8517 
2025-07-10 11:30:51,624 - INFO - machamp.model.callback - morph_accuracy     0.7880   0.9860       0.8179     0.7591 
2025-07-10 11:30:51,624 - INFO - machamp.model.callback - upos_accuracy      0.0645   0.1402       0.9664     0.9124 
2025-07-10 11:30:51,624 - INFO - machamp.model.callback - xpos_accuracy      0.1084   0.1768       0.9426     0.8906 
2025-07-10 11:30:51,624 - INFO - machamp.model.callback - sum                1.6147   4.0791       4.4688     4.0426 
2025-07-10 11:30:51,630 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                           ▗▄▄▄▄▄▄▄▄▄▄▄▞▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄│
│                      ▄▄▀▀▗▄▄▄▄▀▀▀▀▀▀▀▀▀▄▄▄▄▄▄▄▄▄▄▄▄▄▞▀▀▀▀▀▀│
│                    ▄▗▄▄▀▀▘▄▄▄▞▀▀▀▀▀▀▀                      │ 0.8
│                ▗▄▀▀▀▘▄▀▀▀             ▄▄▄▄▄▄▄▛▀▀▀▀▀▀▀▀▀▀▀▀▀│
│              ▄▞▘▘▄▞▀         ▄▄▄▞▀▀▀▀▀                     │
│    ▗▀▀▀▀▀▀▀▀▀▘▄▛▘     ▗▄▄▀▀▀▀▄▄▄▚▄▄▄▄▄▀▚▄▄▀▀▀▞▀▚▄▛▄▞▀▀▀▀▀▀▀│ 0.6
│    ▌      ▗▞▘▘    ▄▄▀▀▘▀▀                                  │
│   ▗▘    ▗▄▀   ▗▄▄▀▀                                        │
│   ▐ ▐▀▗▀▀  ▗▄▗▞▀                                           │
│   ▐ ▌▄▞  ▗▞▘▄▘                                             │ 0.4
│ ▗▞▞▀▀   ▄▀▄▀                                               │
│▄▀ ▌▗▘ ▗▞ ▞                                                 │
│   ▌▐▄▄▄▄▀                                                  │
│  ▐▞▗▘                                                      │ 0.2
│▄▀▞ ▞                                                       │
│ ▗▘▄▘                                                       │
│▀▘▀▘                                                        │
└────────────────────────────────────────────────────────────┘
            10           20           30           40
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:30:51,630 - INFO - machamp.model.trainer - 
2025-07-10 11:30:51,630 - INFO - machamp.model.trainer - Epoch 47/60: training
2025-07-10 11:31:18,132 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:31:18,133 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 6.723466003316751e-06
2025-07-10 11:31:18,133 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 6.723466003316751e-06
2025-07-10 11:31:18,133 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 2.5549170812603656e-06
2025-07-10 11:31:18,133 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:31:18,133 - INFO - machamp.model.trainer - Epoch 47: evaluating on dev
2025-07-10 11:31:19,722 - INFO - machamp.model.callback - epoch       : 47/60
2025-07-10 11:31:19,722 - INFO - machamp.model.callback - best_epoch  : 45
2025-07-10 11:31:19,722 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:31:19,722 - INFO - machamp.model.callback - cur_ram     : 2.1410
2025-07-10 11:31:19,722 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:31:19,722 - INFO - machamp.model.callback - time_total  : 0:23:10
2025-07-10 11:31:19,722 - INFO - machamp.model.callback - 
2025-07-10 11:31:19,722 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:31:19,722 - INFO - machamp.model.callback - Best (45)                                                  
2025-07-10 11:31:19,722 - INFO - machamp.model.callback - dependency_las     0.2317   2.1903       0.8716     0.6276 
2025-07-10 11:31:19,723 - INFO - machamp.model.callback - lemma_accuracy     0.4447   0.5887       0.8673     0.8544 
2025-07-10 11:31:19,723 - INFO - machamp.model.callback - morph_accuracy     0.8084   0.9911       0.8153     0.7615 
2025-07-10 11:31:19,723 - INFO - machamp.model.callback - upos_accuracy      0.0673   0.1389       0.9613     0.9155 
2025-07-10 11:31:19,723 - INFO - machamp.model.callback - xpos_accuracy      0.1107   0.1774       0.9397     0.8940 
2025-07-10 11:31:19,723 - INFO - machamp.model.callback - sum                1.6628   4.0863       4.4553     4.0530 
2025-07-10 11:31:19,723 - INFO - machamp.model.callback - Epoch 47                                                   
2025-07-10 11:31:19,723 - INFO - machamp.model.callback - dependency_las     0.2098   2.2064       0.8771     0.6327 
2025-07-10 11:31:19,723 - INFO - machamp.model.callback - lemma_accuracy     0.4332   0.5847       0.8699     0.8531 
2025-07-10 11:31:19,723 - INFO - machamp.model.callback - morph_accuracy     0.7848   0.9800       0.8192     0.7578 
2025-07-10 11:31:19,723 - INFO - machamp.model.callback - upos_accuracy      0.0647   0.1389       0.9641     0.9118 
2025-07-10 11:31:19,723 - INFO - machamp.model.callback - xpos_accuracy      0.1076   0.1756       0.9438     0.8913 
2025-07-10 11:31:19,723 - INFO - machamp.model.callback - sum                1.6001   4.0855       4.4741     4.0466 
2025-07-10 11:31:19,730 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                           ▄▄▄▄▄▄▄▄▄▄▄▞▗▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄│
│                     ▗▄▞▀▀▄▄▄▄▞▀▀▀▀▀▀▀▀▘▄▄▄▄▄▄▄▄▄▄▄▄▞▀▀▀▀▀▀▀│
│                   ▗▞▄▄▞▀▀▄▄▄▟▀▀▀▀▀▀▀▘                      │ 0.8
│                ▄▞▀▀▀▄▞▀▀▘            ▗▄▄▄▄▄▄▛▀▀▀▀▀▀▀▀▀▀▀▀▀▀│
│             ▗▄▀▀▘▄▀▘        ▄▄▄▄▀▀▀▀▀▘                     │
│    ▄▀▀▀▀▀▀▀▀▘▘▄▀▘     ▄▄▞▀▀▀▄▄▄▞▄▄▄▄▄▞▚▄▄▀▀▀▞▀▚▄▚▄▞▀▀▀▀▀▀▀▀│ 0.6
│    ▌      ▗▞▀     ▗▟▀▀▀▀▘                                  │
│   ▐     ▗▄▘   ▄▗▄▀▘                                        │
│   ▐ ▛▀▗▀▘  ▄▞▄▞▘                                           │
│   ▐ ▌▄▘  ▗▞▗▞                                              │ 0.4
│ ▗▞▌▀▀   ▞▘▞▘                                               │
│▄▀ ▌▐  ▗▛ ▞                                                 │
│   ▌▐▄▄▄▄▀                                                  │
│  ▐▞▗▘                                                      │ 0.2
│▄▀▌ ▞                                                       │
│ ▐ ▞▘                                                       │
│▀▘▀▘                                                        │
└────────────────────────────────────────────────────────────┘
           10           20           30           40
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:31:19,731 - INFO - machamp.model.trainer - 
2025-07-10 11:31:19,731 - INFO - machamp.model.trainer - Epoch 48/60: training
2025-07-10 11:31:46,288 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:31:46,288 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 6.254353233830845e-06
2025-07-10 11:31:46,288 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 6.254353233830845e-06
2025-07-10 11:31:46,289 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 2.3766542288557214e-06
2025-07-10 11:31:46,289 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:31:46,289 - INFO - machamp.model.trainer - Epoch 48: evaluating on dev
2025-07-10 11:31:47,883 - INFO - machamp.model.callback - epoch       : 48/60
2025-07-10 11:31:47,883 - INFO - machamp.model.callback - best_epoch  : 45
2025-07-10 11:31:47,883 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:31:47,883 - INFO - machamp.model.callback - cur_ram     : 2.1410
2025-07-10 11:31:47,883 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:31:47,883 - INFO - machamp.model.callback - time_total  : 0:23:38
2025-07-10 11:31:47,883 - INFO - machamp.model.callback - 
2025-07-10 11:31:47,884 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:31:47,884 - INFO - machamp.model.callback - Best (45)                                                  
2025-07-10 11:31:47,884 - INFO - machamp.model.callback - dependency_las     0.2317   2.1903       0.8716     0.6276 
2025-07-10 11:31:47,884 - INFO - machamp.model.callback - lemma_accuracy     0.4447   0.5887       0.8673     0.8544 
2025-07-10 11:31:47,884 - INFO - machamp.model.callback - morph_accuracy     0.8084   0.9911       0.8153     0.7615 
2025-07-10 11:31:47,884 - INFO - machamp.model.callback - upos_accuracy      0.0673   0.1389       0.9613     0.9155 
2025-07-10 11:31:47,884 - INFO - machamp.model.callback - xpos_accuracy      0.1107   0.1774       0.9397     0.8940 
2025-07-10 11:31:47,884 - INFO - machamp.model.callback - sum                1.6628   4.0863       4.4553     4.0530 
2025-07-10 11:31:47,884 - INFO - machamp.model.callback - Epoch 48                                                   
2025-07-10 11:31:47,884 - INFO - machamp.model.callback - dependency_las     0.1920   2.2086       0.8825     0.6293 
2025-07-10 11:31:47,884 - INFO - machamp.model.callback - lemma_accuracy     0.4256   0.5805       0.8698     0.8544 
2025-07-10 11:31:47,884 - INFO - machamp.model.callback - morph_accuracy     0.7703   0.9721       0.8245     0.7595 
2025-07-10 11:31:47,884 - INFO - machamp.model.callback - upos_accuracy      0.0639   0.1392       0.9641     0.9124 
2025-07-10 11:31:47,884 - INFO - machamp.model.callback - xpos_accuracy      0.1060   0.1756       0.9438     0.8937 
2025-07-10 11:31:47,884 - INFO - machamp.model.callback - sum                1.5578   4.0760       4.4847     4.0493 
2025-07-10 11:31:47,891 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                          ▗▄▄▄▄▄▄▄▄▄▄▄▗▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄│
│                     ▄▄▀▀▗▄▄▄▄▀▀▀▀▀▀▀▀▘▄▄▄▄▄▄▄▄▄▄▄▄▞▀▀▀▀▀▀▀▀│
│                   ▄▀▄▄▀▀▘▄▄▄▀▀▀▀▀▀▀▘                       │ 0.8
│               ▗▄▀▀▀▀▄▀▀▀            ▗▄▄▄▄▄▄▞▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀│
│             ▗▄▘▀▄▞▀        ▗▄▄▄▀▀▀▀▀▘                      │
│    ▄▀▀▀▀▀▀▀▀▘▄▞▀     ▗▄▄▀▀▀▘▄▄▞▄▄▄▄▄▞▚▄▄▀▀▀▟▀▚▄▚▄▞▀▀▀▀▀▀▀▀▀│ 0.6
│   ▗▘      ▗▀▀    ▗▄▞▀▘▀▀                                   │
│   ▐     ▗▞▘   ▄▗▞▀                                         │
│   ▐ ▛▀▐▀▘  ▄▞▄▀▘                                           │
│   ▞ ▌▄▘  ▄▀▄▞                                              │ 0.4
│ ▗▞▌▀▀  ▗▞▗▞                                                │
│▄▀ ▌▐  ▄▘▗▘                                                 │
│  ▗▘▐▄▄▄▄▘                                                  │
│  ▐▞▗▘                                                      │ 0.2
│▄▀▌ ▞                                                       │
│ ▐ ▞▘                                                       │
│▀▘▀▘                                                        │
└────────────────────────────────────────────────────────────┘
           10           20           30          40
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:31:47,891 - INFO - machamp.model.trainer - 
2025-07-10 11:31:47,892 - INFO - machamp.model.trainer - Epoch 49/60: training
2025-07-10 11:32:14,385 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:32:14,386 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 5.785240464344943e-06
2025-07-10 11:32:14,386 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 5.785240464344943e-06
2025-07-10 11:32:14,386 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 2.198391376451078e-06
2025-07-10 11:32:14,386 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:32:14,386 - INFO - machamp.model.trainer - Epoch 49: evaluating on dev
2025-07-10 11:32:15,963 - INFO - machamp.model.callback - epoch       : 49/60
2025-07-10 11:32:15,963 - INFO - machamp.model.callback - best_epoch  : 49
2025-07-10 11:32:15,963 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:32:15,963 - INFO - machamp.model.callback - cur_ram     : 2.1410
2025-07-10 11:32:15,963 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:32:15,963 - INFO - machamp.model.callback - time_total  : 0:24:06
2025-07-10 11:32:15,963 - INFO - machamp.model.callback - 
2025-07-10 11:32:15,963 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:32:15,963 - INFO - machamp.model.callback - Best (49)                                                  
2025-07-10 11:32:15,963 - INFO - machamp.model.callback - dependency_las     0.1846   2.2272       0.8843     0.6344 
2025-07-10 11:32:15,963 - INFO - machamp.model.callback - lemma_accuracy     0.4181   0.5776       0.8732     0.8568 
2025-07-10 11:32:15,963 - INFO - machamp.model.callback - morph_accuracy     0.7582   0.9677       0.8239     0.7608 
2025-07-10 11:32:15,963 - INFO - machamp.model.callback - upos_accuracy      0.0610   0.1395       0.9667     0.9138 
2025-07-10 11:32:15,963 - INFO - machamp.model.callback - xpos_accuracy      0.1044   0.1753       0.9451     0.8953 
2025-07-10 11:32:15,963 - INFO - machamp.model.callback - sum                1.5263   4.0872       4.4933     4.0611 
2025-07-10 11:32:15,963 - INFO - machamp.model.callback - Epoch 49                                                   
2025-07-10 11:32:15,963 - INFO - machamp.model.callback - dependency_las     0.1846   2.2272       0.8843     0.6344 
2025-07-10 11:32:15,963 - INFO - machamp.model.callback - lemma_accuracy     0.4181   0.5776       0.8732     0.8568 
2025-07-10 11:32:15,963 - INFO - machamp.model.callback - morph_accuracy     0.7582   0.9677       0.8239     0.7608 
2025-07-10 11:32:15,964 - INFO - machamp.model.callback - upos_accuracy      0.0610   0.1395       0.9667     0.9138 
2025-07-10 11:32:15,964 - INFO - machamp.model.callback - xpos_accuracy      0.1044   0.1753       0.9451     0.8953 
2025-07-10 11:32:15,964 - INFO - machamp.model.callback - sum                1.5263   4.0872       4.4933     4.0611 
2025-07-10 11:32:15,970 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                          ▄▄▄▄▄▄▄▄▄▄▄▗▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄│
│                     ▄▞▀▀▄▄▄▄▀▀▀▀▀▀▀▀▘▄▄▄▄▄▄▄▄▄▄▄▄▞▀▀▀▀▀▀▀▀▀│
│                  ▗▞▗▄▞▀▀▄▄▄▞▀▀▀▀▀▀▀                        │ 0.8
│               ▗▞▀▀▀▘▞▀▀▘            ▄▄▄▄▄▄▞▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀│
│             ▄▞▘▘▄▀▘        ▄▄▄▞▀▀▀▀▀                       │
│    ▄▀▀▀▀▀▀▀▀▘▄▛▘     ▄▄▞▀▀▀▄▄▄▚▄▄▄▄▞▀▄▄▀▀▀▚▀▚▄▚▄▞▀▀▀▀▀▀▀▀▀▀│ 0.6
│   ▐▘      ▄▘▘    ▗▄▀▀▀▀▘                                   │
│   ▐     ▄▞   ▗▄▄▀▘                                         │
│   ▐ ▛▀▞▀▘ ▗▄▗▞▀                                            │
│   ▌▗▘▄▘  ▄▘▄▘                                              │ 0.4
│ ▗▞▌▀▀  ▗▞▗▀                                                │
│▄▀ ▌▐  ▄▘▗▘                                                 │
│  ▐ ▞▄▄▄▞▘                                                  │
│  ▐▞▐▘                                                      │ 0.2
│▄▀▌ ▌                                                       │
│ ▐ ▞▘                                                       │
│▀▘▀▘                                                        │
└────────────────────────────────────────────────────────────┘
           10          20           30          40
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:32:15,971 - INFO - machamp.model.callback - Performance of 4.0611 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_49.pt
2025-07-10 11:32:17,721 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_45.pt
2025-07-10 11:32:17,848 - INFO - machamp.model.trainer - 
2025-07-10 11:32:17,848 - INFO - machamp.model.trainer - Epoch 50/60: training
2025-07-10 11:32:44,388 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:32:44,388 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 5.316127694859039e-06
2025-07-10 11:32:44,388 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 5.316127694859039e-06
2025-07-10 11:32:44,388 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 2.020128524046435e-06
2025-07-10 11:32:44,388 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:32:44,388 - INFO - machamp.model.trainer - Epoch 50: evaluating on dev
2025-07-10 11:32:46,001 - INFO - machamp.model.callback - epoch       : 50/60
2025-07-10 11:32:46,001 - INFO - machamp.model.callback - best_epoch  : 50
2025-07-10 11:32:46,001 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:32:46,001 - INFO - machamp.model.callback - cur_ram     : 2.1279
2025-07-10 11:32:46,001 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:32:46,001 - INFO - machamp.model.callback - time_total  : 0:24:36
2025-07-10 11:32:46,001 - INFO - machamp.model.callback - 
2025-07-10 11:32:46,001 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:32:46,001 - INFO - machamp.model.callback - Best (50)                                                  
2025-07-10 11:32:46,001 - INFO - machamp.model.callback - dependency_las     0.1905   2.2190       0.8834     0.6347 
2025-07-10 11:32:46,001 - INFO - machamp.model.callback - lemma_accuracy     0.4175   0.5741       0.8739     0.8581 
2025-07-10 11:32:46,001 - INFO - machamp.model.callback - morph_accuracy     0.7546   0.9597       0.8279     0.7642 
2025-07-10 11:32:46,001 - INFO - machamp.model.callback - upos_accuracy      0.0631   0.1382       0.9671     0.9135 
2025-07-10 11:32:46,001 - INFO - machamp.model.callback - xpos_accuracy      0.1040   0.1743       0.9447     0.8950 
2025-07-10 11:32:46,001 - INFO - machamp.model.callback - sum                1.5298   4.0653       4.4970     4.0654 
2025-07-10 11:32:46,001 - INFO - machamp.model.callback - Epoch 50                                                   
2025-07-10 11:32:46,001 - INFO - machamp.model.callback - dependency_las     0.1905   2.2190       0.8834     0.6347 
2025-07-10 11:32:46,002 - INFO - machamp.model.callback - lemma_accuracy     0.4175   0.5741       0.8739     0.8581 
2025-07-10 11:32:46,002 - INFO - machamp.model.callback - morph_accuracy     0.7546   0.9597       0.8279     0.7642 
2025-07-10 11:32:46,002 - INFO - machamp.model.callback - upos_accuracy      0.0631   0.1382       0.9671     0.9135 
2025-07-10 11:32:46,002 - INFO - machamp.model.callback - xpos_accuracy      0.1040   0.1743       0.9447     0.8950 
2025-07-10 11:32:46,002 - INFO - machamp.model.callback - sum                1.5298   4.0653       4.4970     4.0654 
2025-07-10 11:32:46,009 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                         ▗▄▄▄▄▄▄▄▄▄▄▞▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄│
│                    ▗▄▀▀▗▄▄▄▞▀▀▀▀▀▀▀▀▄▄▄▄▄▄▄▄▄▄▄▄▞▀▀▀▀▀▀▀▀▀▀│
│                  ▄▞▄▄▀▀▘▄▄▄▀▀▀▀▀▀▀                         │ 0.8
│               ▄▞▀▀▀▄▀▀▀            ▄▄▄▄▄▄▞▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀│
│             ▄▀▀▘▄▀▘       ▗▄▄▄▀▀▀▀▀                        │
│    ▞▀▀▀▀▀▀▀▀▘▄▀▘    ▗▄▄▀▀▀▘▄▄▞▄▄▄▄▄▀▄▄▞▀▀▚▀▚▄▚▄▞▀▀▀▀▀▀▀▀▀▀▀│ 0.6
│   ▐       ▞▘    ▗▄▞▀▘▀▀                                    │
│   ▐     ▄▛   ▄▗▄▀                                          │
│   ▐ ▛▀▞▀  ▗▄▗▞▘                                            │
│   ▌▐▗▞   ▞▗▞▘                                              │ 0.4
│ ▗▞▌▀▘  ▗▀▗▘                                                │
│▄▀ ▌▐  ▟▘▄▘                                                 │
│  ▐ ▌▄▄▄▞                                                   │
│  ▐▀▐▘                                                      │ 0.2
│▄▀▌▗▌                                                       │
│ ▐ ▞                                                        │
│▀▘▀                                                         │
└────────────────────────────────────────────────────────────┘
           10          20          30          40          50
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:32:46,009 - INFO - machamp.model.callback - Performance of 4.0654 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_50.pt
2025-07-10 11:32:47,799 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_49.pt
2025-07-10 11:32:47,928 - INFO - machamp.model.trainer - 
2025-07-10 11:32:47,928 - INFO - machamp.model.trainer - Epoch 51/60: training
2025-07-10 11:33:14,494 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:33:14,495 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 4.847014925373135e-06
2025-07-10 11:33:14,495 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 4.847014925373135e-06
2025-07-10 11:33:14,495 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 1.8418656716417916e-06
2025-07-10 11:33:14,495 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:33:14,495 - INFO - machamp.model.trainer - Epoch 51: evaluating on dev
2025-07-10 11:33:16,042 - INFO - machamp.model.callback - epoch       : 51/60
2025-07-10 11:33:16,042 - INFO - machamp.model.callback - best_epoch  : 50
2025-07-10 11:33:16,042 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:33:16,042 - INFO - machamp.model.callback - cur_ram     : 2.1366
2025-07-10 11:33:16,042 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:33:16,042 - INFO - machamp.model.callback - time_total  : 0:25:06
2025-07-10 11:33:16,042 - INFO - machamp.model.callback - 
2025-07-10 11:33:16,043 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:33:16,043 - INFO - machamp.model.callback - Best (50)                                                  
2025-07-10 11:33:16,043 - INFO - machamp.model.callback - dependency_las     0.1905   2.2190       0.8834     0.6347 
2025-07-10 11:33:16,043 - INFO - machamp.model.callback - lemma_accuracy     0.4175   0.5741       0.8739     0.8581 
2025-07-10 11:33:16,043 - INFO - machamp.model.callback - morph_accuracy     0.7546   0.9597       0.8279     0.7642 
2025-07-10 11:33:16,043 - INFO - machamp.model.callback - upos_accuracy      0.0631   0.1382       0.9671     0.9135 
2025-07-10 11:33:16,043 - INFO - machamp.model.callback - xpos_accuracy      0.1040   0.1743       0.9447     0.8950 
2025-07-10 11:33:16,043 - INFO - machamp.model.callback - sum                1.5298   4.0653       4.4970     4.0654 
2025-07-10 11:33:16,043 - INFO - machamp.model.callback - Epoch 51                                                   
2025-07-10 11:33:16,043 - INFO - machamp.model.callback - dependency_las     0.1833   2.2279       0.8863     0.6327 
2025-07-10 11:33:16,043 - INFO - machamp.model.callback - lemma_accuracy     0.4107   0.5725       0.8766     0.8574 
2025-07-10 11:33:16,043 - INFO - machamp.model.callback - morph_accuracy     0.7427   0.9546       0.8286     0.7645 
2025-07-10 11:33:16,043 - INFO - machamp.model.callback - upos_accuracy      0.0582   0.1388       0.9672     0.9151 
2025-07-10 11:33:16,043 - INFO - machamp.model.callback - xpos_accuracy      0.1008   0.1748       0.9447     0.8933 
2025-07-10 11:33:16,043 - INFO - machamp.model.callback - sum                1.4958   4.0685       4.5033     4.0631 
2025-07-10 11:33:16,050 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                         ▄▄▄▄▄▄▄▄▄▄▞▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄│
│                    ▄▞▀▀▄▄▄▄▀▀▀▀▀▀▀▀▄▄▄▄▄▄▄▄▄▄▄▄▞▀▀▀▀▀▀▀▀▀▀▀│
│                  ▄▗▄▞▀▀▄▄▄▞▀▀▀▀▀▀▘                         │ 0.8
│               ▄▀▀▀▘▄▀▀▘           ▗▄▄▄▄▄▞▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀│
│            ▗▄▀▀▄▞▀        ▄▄▄▛▀▀▀▀▘                        │
│    ▞▀▀▀▀▀▀▀▘▄▟▀     ▄▄▞▀▀▀▄▄▄▙▄▄▄▄▞▚▄▞▀▀▚▀▚▄▙▄▞▀▀▀▀▀▀▀▀▀▀▀▀│ 0.6
│   ▐      ▗▞▀    ▗▄▀▀▀▀▘                                    │
│   ▐    ▗▄▘   ▄▗▞▀                                          │
│   ▞▗▀▀▛▀  ▗▞▄▀▘                                            │
│   ▌▐▗▞  ▗▞▗▞                                               │ 0.4
│ ▗▀▌▀▘  ▄▘▞▘                                                │
│▄▘▗▘▞ ▗▞ ▞                                                  │
│  ▐ ▗▄▄▄▞                                                   │
│  ▐▀▐▘                                                      │ 0.2
│▄▀▌▐▌                                                       │
│ ▟ ▞                                                        │
│▀▘▀                                                         │
└────────────────────────────────────────────────────────────┘
          10          20          30          40          50
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:33:16,050 - INFO - machamp.model.trainer - 
2025-07-10 11:33:16,050 - INFO - machamp.model.trainer - Epoch 52/60: training
2025-07-10 11:33:42,584 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:33:42,584 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 4.377902155887232e-06
2025-07-10 11:33:42,584 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 4.377902155887232e-06
2025-07-10 11:33:42,584 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 1.6636028192371483e-06
2025-07-10 11:33:42,584 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:33:42,584 - INFO - machamp.model.trainer - Epoch 52: evaluating on dev
2025-07-10 11:33:44,179 - INFO - machamp.model.callback - epoch       : 52/60
2025-07-10 11:33:44,179 - INFO - machamp.model.callback - best_epoch  : 50
2025-07-10 11:33:44,179 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:33:44,179 - INFO - machamp.model.callback - cur_ram     : 2.1369
2025-07-10 11:33:44,179 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:33:44,179 - INFO - machamp.model.callback - time_total  : 0:25:34
2025-07-10 11:33:44,179 - INFO - machamp.model.callback - 
2025-07-10 11:33:44,179 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:33:44,179 - INFO - machamp.model.callback - Best (50)                                                  
2025-07-10 11:33:44,179 - INFO - machamp.model.callback - dependency_las     0.1905   2.2190       0.8834     0.6347 
2025-07-10 11:33:44,180 - INFO - machamp.model.callback - lemma_accuracy     0.4175   0.5741       0.8739     0.8581 
2025-07-10 11:33:44,180 - INFO - machamp.model.callback - morph_accuracy     0.7546   0.9597       0.8279     0.7642 
2025-07-10 11:33:44,180 - INFO - machamp.model.callback - upos_accuracy      0.0631   0.1382       0.9671     0.9135 
2025-07-10 11:33:44,180 - INFO - machamp.model.callback - xpos_accuracy      0.1040   0.1743       0.9447     0.8950 
2025-07-10 11:33:44,180 - INFO - machamp.model.callback - sum                1.5298   4.0653       4.4970     4.0654 
2025-07-10 11:33:44,180 - INFO - machamp.model.callback - Epoch 52                                                   
2025-07-10 11:33:44,180 - INFO - machamp.model.callback - dependency_las     0.1748   2.2455       0.8869     0.6320 
2025-07-10 11:33:44,180 - INFO - machamp.model.callback - lemma_accuracy     0.4072   0.5706       0.8759     0.8588 
2025-07-10 11:33:44,180 - INFO - machamp.model.callback - morph_accuracy     0.7325   0.9505       0.8290     0.7642 
2025-07-10 11:33:44,180 - INFO - machamp.model.callback - upos_accuracy      0.0584   0.1382       0.9674     0.9131 
2025-07-10 11:33:44,180 - INFO - machamp.model.callback - xpos_accuracy      0.0991   0.1735       0.9486     0.8967 
2025-07-10 11:33:44,180 - INFO - machamp.model.callback - sum                1.4720   4.0782       4.5077     4.0647 
2025-07-10 11:33:44,187 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                        ▗▄▄▄▄▄▄▄▄▄▄▗▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄│
│                   ▗▄▀▀▗▄▄▄▞▀▀▀▀▀▀▀▘▄▄▄▄▄▄▄▄▄▄▄▞▀▀▀▀▀▀▀▀▀▀▀▀│
│                 ▗▞▄▄▀▀▘▄▄▄▀▀▀▀▀▀▀                          │ 0.8
│              ▗▞▀▀▀▄▞▀▀           ▗▄▄▄▄▄▄▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀│
│            ▗▞▘▘▄▀▘       ▗▄▄▞▀▀▀▀▘                         │
│    ▞▀▀▀▀▀▀▀▘▄▞▘    ▗▄▄▀▀▀▘▄▄▙▄▄▄▄▞▚▄▄▀▀▀▛▚▄▞▄▞▀▀▀▀▀▀▀▀▀▀▀▀▀│ 0.6
│   ▐      ▗▀▘    ▗▞▀▘▀▀                                     │
│   ▐    ▗▟▘  ▗▄▄▀▘                                          │
│   ▌▐▀▀▛▘  ▄▞▄▀                                             │
│   ▌▐▗▞  ▗▀▗▞                                               │ 0.4
│ ▗▀▌▀▘  ▄▘▞▘                                                │
│▄▘▐ ▌ ▗▞ ▞                                                  │
│  ▐ ▗▄▄▄▀                                                   │
│  ▐▀▞                                                       │ 0.2
│▄▀▌▗▘                                                       │
│ ▟ ▞                                                        │
│▀▘▀                                                         │
└────────────────────────────────────────────────────────────┘
          10          20          30         40          50
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:33:44,187 - INFO - machamp.model.trainer - 
2025-07-10 11:33:44,187 - INFO - machamp.model.trainer - Epoch 53/60: training
2025-07-10 11:34:10,685 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:34:10,685 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 3.908789386401326e-06
2025-07-10 11:34:10,685 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 3.908789386401326e-06
2025-07-10 11:34:10,685 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 1.485339966832504e-06
2025-07-10 11:34:10,685 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:34:10,686 - INFO - machamp.model.trainer - Epoch 53: evaluating on dev
2025-07-10 11:34:12,253 - INFO - machamp.model.callback - epoch       : 53/60
2025-07-10 11:34:12,253 - INFO - machamp.model.callback - best_epoch  : 53
2025-07-10 11:34:12,253 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:34:12,253 - INFO - machamp.model.callback - cur_ram     : 2.1369
2025-07-10 11:34:12,253 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:34:12,253 - INFO - machamp.model.callback - time_total  : 0:26:02
2025-07-10 11:34:12,254 - INFO - machamp.model.callback - 
2025-07-10 11:34:12,254 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:34:12,254 - INFO - machamp.model.callback - Best (53)                                                  
2025-07-10 11:34:12,254 - INFO - machamp.model.callback - dependency_las     0.1684   2.2465       0.8895     0.6354 
2025-07-10 11:34:12,254 - INFO - machamp.model.callback - lemma_accuracy     0.4046   0.5682       0.8744     0.8581 
2025-07-10 11:34:12,254 - INFO - machamp.model.callback - morph_accuracy     0.7218   0.9456       0.8328     0.7659 
2025-07-10 11:34:12,254 - INFO - machamp.model.callback - upos_accuracy      0.0564   0.1385       0.9696     0.9131 
2025-07-10 11:34:12,254 - INFO - machamp.model.callback - xpos_accuracy      0.0973   0.1742       0.9492     0.8963 
2025-07-10 11:34:12,254 - INFO - machamp.model.callback - sum                1.4484   4.0731       4.5155     4.0688 
2025-07-10 11:34:12,254 - INFO - machamp.model.callback - Epoch 53                                                   
2025-07-10 11:34:12,255 - INFO - machamp.model.callback - dependency_las     0.1684   2.2465       0.8895     0.6354 
2025-07-10 11:34:12,255 - INFO - machamp.model.callback - lemma_accuracy     0.4046   0.5682       0.8744     0.8581 
2025-07-10 11:34:12,255 - INFO - machamp.model.callback - morph_accuracy     0.7218   0.9456       0.8328     0.7659 
2025-07-10 11:34:12,255 - INFO - machamp.model.callback - upos_accuracy      0.0564   0.1385       0.9696     0.9131 
2025-07-10 11:34:12,255 - INFO - machamp.model.callback - xpos_accuracy      0.0973   0.1742       0.9492     0.8963 
2025-07-10 11:34:12,255 - INFO - machamp.model.callback - sum                1.4484   4.0731       4.5155     4.0688 
2025-07-10 11:34:12,262 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                        ▄▄▄▄▄▄▄▄▄▄▞▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄│
│                   ▄▞▀▀▄▄▄▄▀▀▀▀▀▀▀▀▄▄▄▄▄▄▄▄▄▄▄▞▀▀▀▀▀▀▀▀▀▀▀▀▀│
│                 ▄▀▄▞▀▀▄▄▄▞▀▀▀▀▀▀                          ▄│ 0.8
│              ▄▞▀▀▀▄▀▀▘           ▄▄▄▄▄▄▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀ │
│            ▄▞▘▘▄▀        ▄▄▄▀▀▀▀▀                          │
│    ▞▀▀▀▀▀▀▀▘▄▛▘    ▄▄▞▀▀▀▄▄▞▄▄▄▄▄▀▄▄▀▀▀▞▀▄▞▄▞▀▀▀▀▀▀▀▀▀▀▀▀▀▀│ 0.6
│   ▐      ▗▘▘   ▗▄▞▀▀▀▘                                     │
│   ▞    ▗▞▘  ▗▄▄▀                                           │
│   ▌▐▀▗▀▘  ▄▗▞▀                                             │
│   ▌▐▗▞  ▄▀▄▘                                               │ 0.4
│ ▗▀▌▀▘  ▞▗▀                                                 │
│▞▘▐ ▌ ▗▞▗▘                                                  │
│  ▐ ▗▄▄▄▘                                                   │
│  ▞▗▞                                                       │ 0.2
│▄▀▌▗▘                                                       │
│ ▟▗▞                                                        │
│▀▘▀                                                         │
└────────────────────────────────────────────────────────────┘
          10         20          30         40          50
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:34:12,262 - INFO - machamp.model.callback - Performance of 4.0688 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_53.pt
2025-07-10 11:34:14,066 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_50.pt
2025-07-10 11:34:14,198 - INFO - machamp.model.trainer - 
2025-07-10 11:34:14,198 - INFO - machamp.model.trainer - Epoch 54/60: training
2025-07-10 11:34:40,680 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:34:40,680 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 3.4396766169154226e-06
2025-07-10 11:34:40,680 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 3.4396766169154226e-06
2025-07-10 11:34:40,680 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 1.3070771144278606e-06
2025-07-10 11:34:40,680 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:34:40,681 - INFO - machamp.model.trainer - Epoch 54: evaluating on dev
2025-07-10 11:34:42,248 - INFO - machamp.model.callback - epoch       : 54/60
2025-07-10 11:34:42,248 - INFO - machamp.model.callback - best_epoch  : 54
2025-07-10 11:34:42,248 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:34:42,248 - INFO - machamp.model.callback - cur_ram     : 2.1410
2025-07-10 11:34:42,248 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:34:42,248 - INFO - machamp.model.callback - time_total  : 0:26:32
2025-07-10 11:34:42,249 - INFO - machamp.model.callback - 
2025-07-10 11:34:42,249 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:34:42,249 - INFO - machamp.model.callback - Best (54)                                                  
2025-07-10 11:34:42,249 - INFO - machamp.model.callback - dependency_las     0.1778   2.2441       0.8889     0.6333 
2025-07-10 11:34:42,249 - INFO - machamp.model.callback - lemma_accuracy     0.3999   0.5673       0.8785     0.8601 
2025-07-10 11:34:42,249 - INFO - machamp.model.callback - morph_accuracy     0.7171   0.9441       0.8340     0.7669 
2025-07-10 11:34:42,249 - INFO - machamp.model.callback - upos_accuracy      0.0566   0.1389       0.9681     0.9131 
2025-07-10 11:34:42,249 - INFO - machamp.model.callback - xpos_accuracy      0.0974   0.1742       0.9478     0.8963 
2025-07-10 11:34:42,249 - INFO - machamp.model.callback - sum                1.4488   4.0686       4.5174     4.0698 
2025-07-10 11:34:42,249 - INFO - machamp.model.callback - Epoch 54                                                   
2025-07-10 11:34:42,249 - INFO - machamp.model.callback - dependency_las     0.1778   2.2441       0.8889     0.6333 
2025-07-10 11:34:42,249 - INFO - machamp.model.callback - lemma_accuracy     0.3999   0.5673       0.8785     0.8601 
2025-07-10 11:34:42,249 - INFO - machamp.model.callback - morph_accuracy     0.7171   0.9441       0.8340     0.7669 
2025-07-10 11:34:42,249 - INFO - machamp.model.callback - upos_accuracy      0.0566   0.1389       0.9681     0.9131 
2025-07-10 11:34:42,249 - INFO - machamp.model.callback - xpos_accuracy      0.0974   0.1742       0.9478     0.8963 
2025-07-10 11:34:42,249 - INFO - machamp.model.callback - sum                1.4488   4.0686       4.5174     4.0698 
2025-07-10 11:34:42,256 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                       ▗▄▄▄▄▄▄▄▄▄▞▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄│
│                   ▄▛▀▗▄▄▄▞▀▀▀▀▀▀▀▄▄▄▄▄▄▄▄▄▄▄▞▀▀▀▀▀▀▀▀▀▀▀▀▀▀│
│                 ▄▗▄▀▀▘▄▄▄▀▀▀▀▀▀▘                         ▄▄│ 0.8
│              ▄▀▀▀▘▞▀▀           ▗▄▄▄▄▄▞▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀  │
│            ▄▀▀▄▞▀       ▗▄▄▞▀▀▀▀▘                          │
│   ▗▞▀▀▀▀▀▀▀▘▄▀     ▄▄▞▀▀▘▄▄▚▄▄▄▄▞▄▄▞▀▀▟▀▄▞▄▄▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀│ 0.6
│   ▐      ▄▘▘   ▗▄▀▀▀▀▘                                     │
│   ▌    ▗▞   ▄▗▞▀                                           │
│   ▌▐▀▗▀▘ ▗▄▗▞▘                                             │
│   ▌▐▗▌  ▄▘▄▘                                               │ 0.4
│ ▗▗▘▀▘  ▞▗▀                                                 │
│▞▘▐ ▌ ▗▛▗▘                                                  │
│  ▐ ▗▄▄▄▘                                                   │
│  ▞▐▌                                                       │ 0.2
│▄▗▘▗▘                                                       │
│ ▟▗▞                                                        │
│▀▘▀                                                         │
└────────────────────────────────────────────────────────────┘
          10         20         30          40         50
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:34:42,257 - INFO - machamp.model.callback - Performance of 4.0698 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_54.pt
2025-07-10 11:34:44,040 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_53.pt
2025-07-10 11:34:44,170 - INFO - machamp.model.trainer - 
2025-07-10 11:34:44,171 - INFO - machamp.model.trainer - Epoch 55/60: training
2025-07-10 11:35:10,724 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:35:10,724 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 2.9705638474295196e-06
2025-07-10 11:35:10,724 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 2.9705638474295196e-06
2025-07-10 11:35:10,724 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 1.1288142620232176e-06
2025-07-10 11:35:10,724 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:35:10,724 - INFO - machamp.model.trainer - Epoch 55: evaluating on dev
2025-07-10 11:35:12,327 - INFO - machamp.model.callback - epoch       : 55/60
2025-07-10 11:35:12,327 - INFO - machamp.model.callback - best_epoch  : 55
2025-07-10 11:35:12,327 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:35:12,327 - INFO - machamp.model.callback - cur_ram     : 2.1410
2025-07-10 11:35:12,327 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:35:12,327 - INFO - machamp.model.callback - time_total  : 0:27:02
2025-07-10 11:35:12,327 - INFO - machamp.model.callback - 
2025-07-10 11:35:12,327 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:35:12,327 - INFO - machamp.model.callback - Best (55)                                                  
2025-07-10 11:35:12,327 - INFO - machamp.model.callback - dependency_las     0.1582   2.2490       0.8927     0.6354 
2025-07-10 11:35:12,327 - INFO - machamp.model.callback - lemma_accuracy     0.3954   0.5669       0.8780     0.8604 
2025-07-10 11:35:12,327 - INFO - machamp.model.callback - morph_accuracy     0.7103   0.9422       0.8362     0.7662 
2025-07-10 11:35:12,328 - INFO - machamp.model.callback - upos_accuracy      0.0566   0.1388       0.9693     0.9145 
2025-07-10 11:35:12,328 - INFO - machamp.model.callback - xpos_accuracy      0.0968   0.1739       0.9484     0.8967 
2025-07-10 11:35:12,328 - INFO - machamp.model.callback - sum                1.4173   4.0708       4.5246     4.0731 
2025-07-10 11:35:12,328 - INFO - machamp.model.callback - Epoch 55                                                   
2025-07-10 11:35:12,328 - INFO - machamp.model.callback - dependency_las     0.1582   2.2490       0.8927     0.6354 
2025-07-10 11:35:12,328 - INFO - machamp.model.callback - lemma_accuracy     0.3954   0.5669       0.8780     0.8604 
2025-07-10 11:35:12,328 - INFO - machamp.model.callback - morph_accuracy     0.7103   0.9422       0.8362     0.7662 
2025-07-10 11:35:12,328 - INFO - machamp.model.callback - upos_accuracy      0.0566   0.1388       0.9693     0.9145 
2025-07-10 11:35:12,328 - INFO - machamp.model.callback - xpos_accuracy      0.0968   0.1739       0.9484     0.8967 
2025-07-10 11:35:12,328 - INFO - machamp.model.callback - sum                1.4173   4.0708       4.5246     4.0731 
2025-07-10 11:35:12,334 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                       ▄▄▄▄▄▄▄▄▄▄▗▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄│
│                  ▗▄▀▀▄▄▄▄▀▀▀▀▀▀▀▘▄▄▄▄▄▄▄▄▄▄▄▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀│
│                ▗▞▄▄▀▀▄▄▄▞▀▀▀▀▀▀                         ▄▄▄│ 0.8
│             ▗▄▀▀▀▄▞▀▀          ▗▄▄▄▄▄▞▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀   │
│           ▗▄▀▀▄▀▘       ▄▄▄▀▀▀▀▘                           │
│   ▗▞▀▀▀▀▀▀▘▄▟▀    ▗▄▄▀▀▀▄▄▞▄▄▄▄▞▚▄▞▀▀▚▀▚▄▚▄▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀│ 0.6
│   ▐      ▞▀    ▗▞▀▘▀▀                                      │
│   ▌    ▄▛   ▄▄▞▘                                           │
│   ▌▐▀▐▀▘ ▗▞▄▀                                              │
│   ▌▞▄▘  ▞▗▞                                                │ 0.4
│ ▗▐▀▀  ▗▞▗▘                                                 │
│▞▘▐ ▌ ▄▘▗▘                                                  │
│  ▐▗▗▄▄▞▘                                                   │
│  ▞▐▌                                                       │ 0.2
│▄▗▘▐                                                        │
│ ▟▗▀                                                        │
│▀▘▀                                                         │
└────────────────────────────────────────────────────────────┘
          10         20         30         40         50
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:35:12,335 - INFO - machamp.model.callback - Performance of 4.0731 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_55.pt
2025-07-10 11:35:14,132 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_54.pt
2025-07-10 11:35:14,268 - INFO - machamp.model.trainer - 
2025-07-10 11:35:14,268 - INFO - machamp.model.trainer - Epoch 56/60: training
2025-07-10 11:35:40,851 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:35:40,852 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 2.5014510779436163e-06
2025-07-10 11:35:40,852 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 2.5014510779436163e-06
2025-07-10 11:35:40,852 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 9.505514096185743e-07
2025-07-10 11:35:40,852 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:35:40,852 - INFO - machamp.model.trainer - Epoch 56: evaluating on dev
2025-07-10 11:35:42,447 - INFO - machamp.model.callback - epoch       : 56/60
2025-07-10 11:35:42,447 - INFO - machamp.model.callback - best_epoch  : 56
2025-07-10 11:35:42,447 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:35:42,447 - INFO - machamp.model.callback - cur_ram     : 2.1411
2025-07-10 11:35:42,447 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:35:42,447 - INFO - machamp.model.callback - time_total  : 0:27:33
2025-07-10 11:35:42,447 - INFO - machamp.model.callback - 
2025-07-10 11:35:42,447 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:35:42,447 - INFO - machamp.model.callback - Best (56)                                                  
2025-07-10 11:35:42,447 - INFO - machamp.model.callback - dependency_las     0.1559   2.2455       0.8949     0.6354 
2025-07-10 11:35:42,448 - INFO - machamp.model.callback - lemma_accuracy     0.3950   0.5657       0.8802     0.8598 
2025-07-10 11:35:42,448 - INFO - machamp.model.callback - morph_accuracy     0.7065   0.9405       0.8326     0.7685 
2025-07-10 11:35:42,448 - INFO - machamp.model.callback - upos_accuracy      0.0553   0.1385       0.9698     0.9145 
2025-07-10 11:35:42,448 - INFO - machamp.model.callback - xpos_accuracy      0.0952   0.1738       0.9506     0.8984 
2025-07-10 11:35:42,448 - INFO - machamp.model.callback - sum                1.4079   4.0639       4.5281     4.0765 
2025-07-10 11:35:42,448 - INFO - machamp.model.callback - Epoch 56                                                   
2025-07-10 11:35:42,448 - INFO - machamp.model.callback - dependency_las     0.1559   2.2455       0.8949     0.6354 
2025-07-10 11:35:42,448 - INFO - machamp.model.callback - lemma_accuracy     0.3950   0.5657       0.8802     0.8598 
2025-07-10 11:35:42,448 - INFO - machamp.model.callback - morph_accuracy     0.7065   0.9405       0.8326     0.7685 
2025-07-10 11:35:42,448 - INFO - machamp.model.callback - upos_accuracy      0.0553   0.1385       0.9698     0.9145 
2025-07-10 11:35:42,448 - INFO - machamp.model.callback - xpos_accuracy      0.0952   0.1738       0.9506     0.8984 
2025-07-10 11:35:42,448 - INFO - machamp.model.callback - sum                1.4079   4.0639       4.5281     4.0765 
2025-07-10 11:35:42,455 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                      ▗▄▄▄▄▄▄▄▄▄▞▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄│
│                  ▄▞▀▗▄▄▄▞▀▀▀▀▀▀▀▄▄▄▄▄▄▄▄▄▄▄▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀│
│                ▗▀▄▞▀▘▄▄▄▀▀▀▀▀▀▘                        ▄▄▄▄│ 0.8
│             ▗▞▀▀▀▄▀▀▘          ▄▄▄▄▄▄▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀    │
│           ▗▞▘▘▄▀       ▗▄▄▞▀▀▀▀                            │
│   ▗▀▀▀▀▀▀▀▘▄▞▘    ▄▄▞▀▀▘▄▄▚▄▄▄▄▀▄▄▀▀▀▛▚▄▚▄▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀│ 0.6
│   ▞     ▗▞▀   ▗▗▞▀▀▀▘                                      │
│   ▌    ▄▘  ▗▄▄▀▘                                           │
│   ▌▐▀▐▀  ▗▞▄▀                                              │
│   ▌▌▄▘ ▗▞▗▞                                                │ 0.4
│ ▗▐▀▀  ▗▀▞▘                                                 │
│▞▘▐ ▌ ▄▘▐                                                   │
│  ▐▐▗▄▄▞▘                                                   │
│  ▛▐▌                                                       │ 0.2
│▄▗▘▐                                                        │
│ ▞▗▀                                                        │
│▀▘▘                                                         │
└────────────────────────────────────────────────────────────┘
         10         20         30         40         50
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:35:42,456 - INFO - machamp.model.callback - Performance of 4.0765 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_56.pt
2025-07-10 11:35:44,313 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_55.pt
2025-07-10 11:35:44,447 - INFO - machamp.model.trainer - 
2025-07-10 11:35:44,448 - INFO - machamp.model.trainer - Epoch 57/60: training
2025-07-10 11:36:11,028 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:36:11,028 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 2.0323383084577125e-06
2025-07-10 11:36:11,028 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 2.0323383084577125e-06
2025-07-10 11:36:11,028 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 7.722885572139308e-07
2025-07-10 11:36:11,028 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:36:11,029 - INFO - machamp.model.trainer - Epoch 57: evaluating on dev
2025-07-10 11:36:12,600 - INFO - machamp.model.callback - epoch       : 57/60
2025-07-10 11:36:12,600 - INFO - machamp.model.callback - best_epoch  : 57
2025-07-10 11:36:12,600 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:36:12,600 - INFO - machamp.model.callback - cur_ram     : 2.1411
2025-07-10 11:36:12,600 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:36:12,600 - INFO - machamp.model.callback - time_total  : 0:28:03
2025-07-10 11:36:12,600 - INFO - machamp.model.callback - 
2025-07-10 11:36:12,601 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:36:12,601 - INFO - machamp.model.callback - Best (57)                                                  
2025-07-10 11:36:12,601 - INFO - machamp.model.callback - dependency_las     0.1592   2.2479       0.8914     0.6380 
2025-07-10 11:36:12,601 - INFO - machamp.model.callback - lemma_accuracy     0.3926   0.5634       0.8807     0.8608 
2025-07-10 11:36:12,601 - INFO - machamp.model.callback - morph_accuracy     0.7044   0.9371       0.8350     0.7675 
2025-07-10 11:36:12,601 - INFO - machamp.model.callback - upos_accuracy      0.0547   0.1386       0.9707     0.9151 
2025-07-10 11:36:12,601 - INFO - machamp.model.callback - xpos_accuracy      0.0921   0.1734       0.9522     0.8984 
2025-07-10 11:36:12,601 - INFO - machamp.model.callback - sum                1.4031   4.0604       4.5299     4.0798 
2025-07-10 11:36:12,601 - INFO - machamp.model.callback - Epoch 57                                                   
2025-07-10 11:36:12,601 - INFO - machamp.model.callback - dependency_las     0.1592   2.2479       0.8914     0.6380 
2025-07-10 11:36:12,601 - INFO - machamp.model.callback - lemma_accuracy     0.3926   0.5634       0.8807     0.8608 
2025-07-10 11:36:12,601 - INFO - machamp.model.callback - morph_accuracy     0.7044   0.9371       0.8350     0.7675 
2025-07-10 11:36:12,601 - INFO - machamp.model.callback - upos_accuracy      0.0547   0.1386       0.9707     0.9151 
2025-07-10 11:36:12,601 - INFO - machamp.model.callback - xpos_accuracy      0.0921   0.1734       0.9522     0.8984 
2025-07-10 11:36:12,601 - INFO - machamp.model.callback - sum                1.4031   4.0604       4.5299     4.0798 
2025-07-10 11:36:12,608 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                      ▗▄▄▄▄▄▄▄▄▄▗▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄│
│                  ▄▛▀▗▄▄▄▀▀▀▀▀▀▀▘▄▄▄▄▄▄▄▄▄▄▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀│
│                ▄▗▄▀▀▘▄▄▞▀▀▀▀▀▀                        ▄▄▄▄▄│ 0.8
│             ▄▞▀▀▘▞▀▀          ▗▄▄▄▄▄▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀     │
│           ▄▞▘▘▞▀       ▄▄▄▀▀▀▀▘                            │
│   ▗▀▀▀▀▀▀▀▞▄▛▘    ▄▄▀▀▀▄▄▞▄▄▄▄▞▚▄▞▀▀▞▀▄▞▄▞▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀│ 0.6
│   ▌     ▗▀▘   ▗▄▀▀▀▀                                       │
│   ▌    ▄▘  ▗▗▞▀                                            │
│   ▌▞▀▞▀  ▄▗▞▘                                              │
│  ▗▘▌▄▘ ▗▞▗▘                                                │ 0.4
│ ▗▐▀▀  ▗▘▞▘                                                 │
│▞▘▐ ▌ ▄▘▞                                                   │
│  ▐▐▄▄▄▞                                                    │
│  ▌▐▌                                                       │ 0.2
│▄▗▘▐                                                        │
│ ▞▗▘                                                        │
│▀▘▘                                                         │
└────────────────────────────────────────────────────────────┘
         10         20         30        40         50
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:36:12,608 - INFO - machamp.model.callback - Performance of 4.0798 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_57.pt
2025-07-10 11:36:14,405 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_56.pt
2025-07-10 11:36:14,538 - INFO - machamp.model.trainer - 
2025-07-10 11:36:14,538 - INFO - machamp.model.trainer - Epoch 58/60: training
2025-07-10 11:36:41,147 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:36:41,148 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.563225538971807e-06
2025-07-10 11:36:41,148 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.563225538971807e-06
2025-07-10 11:36:41,148 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 5.940257048092867e-07
2025-07-10 11:36:41,148 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:36:41,148 - INFO - machamp.model.trainer - Epoch 58: evaluating on dev
2025-07-10 11:36:42,754 - INFO - machamp.model.callback - epoch       : 58/60
2025-07-10 11:36:42,754 - INFO - machamp.model.callback - best_epoch  : 57
2025-07-10 11:36:42,754 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:36:42,754 - INFO - machamp.model.callback - cur_ram     : 2.1281
2025-07-10 11:36:42,754 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:36:42,754 - INFO - machamp.model.callback - time_total  : 0:28:33
2025-07-10 11:36:42,754 - INFO - machamp.model.callback - 
2025-07-10 11:36:42,754 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:36:42,754 - INFO - machamp.model.callback - Best (57)                                                  
2025-07-10 11:36:42,754 - INFO - machamp.model.callback - dependency_las     0.1592   2.2479       0.8914     0.6380 
2025-07-10 11:36:42,754 - INFO - machamp.model.callback - lemma_accuracy     0.3926   0.5634       0.8807     0.8608 
2025-07-10 11:36:42,754 - INFO - machamp.model.callback - morph_accuracy     0.7044   0.9371       0.8350     0.7675 
2025-07-10 11:36:42,754 - INFO - machamp.model.callback - upos_accuracy      0.0547   0.1386       0.9707     0.9151 
2025-07-10 11:36:42,754 - INFO - machamp.model.callback - xpos_accuracy      0.0921   0.1734       0.9522     0.8984 
2025-07-10 11:36:42,754 - INFO - machamp.model.callback - sum                1.4031   4.0604       4.5299     4.0798 
2025-07-10 11:36:42,754 - INFO - machamp.model.callback - Epoch 58                                                   
2025-07-10 11:36:42,755 - INFO - machamp.model.callback - dependency_las     0.1554   2.2586       0.8924     0.6360 
2025-07-10 11:36:42,755 - INFO - machamp.model.callback - lemma_accuracy     0.3921   0.5623       0.8799     0.8611 
2025-07-10 11:36:42,755 - INFO - machamp.model.callback - morph_accuracy     0.6985   0.9350       0.8360     0.7682 
2025-07-10 11:36:42,755 - INFO - machamp.model.callback - upos_accuracy      0.0548   0.1384       0.9702     0.9151 
2025-07-10 11:36:42,755 - INFO - machamp.model.callback - xpos_accuracy      0.0930   0.1733       0.9506     0.8980 
2025-07-10 11:36:42,755 - INFO - machamp.model.callback - sum                1.3937   4.0676       4.5291     4.0785 
2025-07-10 11:36:42,762 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                      ▄▄▄▄▄▄▄▄▄▞▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄│
│                 ▗▄▀▀▄▄▄▞▀▀▀▀▀▀▀▄▄▄▄▄▄▄▄▄▄▞▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀│
│               ▗▞▄▄▀▀▄▄▄▀▀▀▀▀▀                        ▄▄▄▄▄▄│ 0.8
│             ▄▀▀▀▄▞▀▘          ▄▄▄▄▄▞▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀      │
│           ▄▀▀▄▞▘      ▗▄▄▞▀▀▀▀                             │
│   ▗▀▀▀▀▀▀▀▘▄▀    ▗▄▞▀▀▘▄▄▚▄▄▄▄▀▄▄▀▀▚▀▚▟▄▞▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀│ 0.6
│   ▌     ▗▘▘   ▗▟▀▘▀▘                                       │
│   ▌   ▗▟▘  ▗▗▞▘                                            │
│   ▌▛▀▞▀  ▄▗▞▘                                              │
│  ▐ ▌▟  ▗▀▄▘                                                │ 0.4
│ ▗▐▀▀  ▗▘▞                                                  │
│▞▘▐▗▘ ▟▘▞                                                   │
│  ▞▐▄▄▄▞                                                    │
│  ▌▗▘                                                       │ 0.2
│▞▗▘▐                                                        │
│ ▞▗▘                                                        │
│▀▘▘                                                         │
└────────────────────────────────────────────────────────────┘
         10         20        30         40        50
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:36:42,762 - INFO - machamp.model.trainer - 
2025-07-10 11:36:42,762 - INFO - machamp.model.trainer - Epoch 59/60: training
2025-07-10 11:37:09,280 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:37:09,280 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 1.0941127694859035e-06
2025-07-10 11:37:09,281 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 1.0941127694859035e-06
2025-07-10 11:37:09,281 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 4.1576285240464334e-07
2025-07-10 11:37:09,281 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:37:09,281 - INFO - machamp.model.trainer - Epoch 59: evaluating on dev
2025-07-10 11:37:10,847 - INFO - machamp.model.callback - epoch       : 59/60
2025-07-10 11:37:10,847 - INFO - machamp.model.callback - best_epoch  : 59
2025-07-10 11:37:10,847 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:37:10,847 - INFO - machamp.model.callback - cur_ram     : 2.1281
2025-07-10 11:37:10,847 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:37:10,847 - INFO - machamp.model.callback - time_total  : 0:29:01
2025-07-10 11:37:10,847 - INFO - machamp.model.callback - 
2025-07-10 11:37:10,847 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:37:10,847 - INFO - machamp.model.callback - Best (59)                                                  
2025-07-10 11:37:10,847 - INFO - machamp.model.callback - dependency_las     0.1509   2.2613       0.8962     0.6374 
2025-07-10 11:37:10,847 - INFO - machamp.model.callback - lemma_accuracy     0.3895   0.5621       0.8825     0.8618 
2025-07-10 11:37:10,847 - INFO - machamp.model.callback - morph_accuracy     0.6977   0.9341       0.8375     0.7689 
2025-07-10 11:37:10,848 - INFO - machamp.model.callback - upos_accuracy      0.0552   0.1384       0.9703     0.9141 
2025-07-10 11:37:10,848 - INFO - machamp.model.callback - xpos_accuracy      0.0942   0.1735       0.9519     0.8984 
2025-07-10 11:37:10,848 - INFO - machamp.model.callback - sum                1.3874   4.0695       4.5385     4.0805 
2025-07-10 11:37:10,848 - INFO - machamp.model.callback - Epoch 59                                                   
2025-07-10 11:37:10,848 - INFO - machamp.model.callback - dependency_las     0.1509   2.2613       0.8962     0.6374 
2025-07-10 11:37:10,848 - INFO - machamp.model.callback - lemma_accuracy     0.3895   0.5621       0.8825     0.8618 
2025-07-10 11:37:10,848 - INFO - machamp.model.callback - morph_accuracy     0.6977   0.9341       0.8375     0.7689 
2025-07-10 11:37:10,848 - INFO - machamp.model.callback - upos_accuracy      0.0552   0.1384       0.9703     0.9141 
2025-07-10 11:37:10,848 - INFO - machamp.model.callback - xpos_accuracy      0.0942   0.1735       0.9519     0.8984 
2025-07-10 11:37:10,848 - INFO - machamp.model.callback - sum                1.3874   4.0695       4.5385     4.0805 
2025-07-10 11:37:10,856 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                     ▗▄▄▄▄▄▄▄▄▄▗▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄│
│                 ▗▞▀▗▄▄▄▀▀▀▀▀▀▀▘▄▄▄▄▄▄▄▄▄▄▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀│
│               ▗▞▄▞▀▘▄▄▞▀▀▀▀▀▘                       ▄▄▄▄▄▄▄│ 0.8
│             ▄▀▀▀▄▀▀▘         ▗▄▄▄▄▄▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀       │
│           ▄▀▀▄▀▘      ▗▄▄▀▀▀▀▘                             │
│   ▗▀▀▀▀▀▀▀▘▟▀    ▄▄▞▀▀▘▄▟▄▄▄▄▞▚▄▞▀▀▞▚▄▚▄▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀│ 0.6
│   ▌     ▗▘    ▗▞▀▀▀▘                                       │
│   ▌   ▗▞▘  ▄▗▀▘                                            │
│   ▌▛▀▛▘  ▄▗▀▘                                              │
│  ▐ ▌▟  ▄▀▄▘                                                │ 0.4
│ ▗▐▀▀  ▄▗▀                                                  │
│▞▘▐▐  ▞ ▛                                                   │
│  ▞▐▄▄▄▀                                                    │
│  ▌▗▘                                                       │ 0.2
│▞▗▘▐                                                        │
│ ▞▗▘                                                        │
│▀▘▘                                                         │
└────────────────────────────────────────────────────────────┘
         10        20         30        40        50
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:37:10,856 - INFO - machamp.model.callback - Performance of 4.0805 within top 1 models, saving to logs/ota_config/2025.07.10_11.07.44/model_59.pt
2025-07-10 11:37:12,635 - INFO - machamp.model.callback - Removing old model: logs/ota_config/2025.07.10_11.07.44/model_57.pt
2025-07-10 11:37:12,765 - INFO - machamp.model.trainer - 
2025-07-10 11:37:12,765 - INFO - machamp.model.trainer - Epoch 60/60: training
2025-07-10 11:37:39,262 - INFO - machamp.modules.allennlp.slanted_triangular - Learning rates for each group: 

2025-07-10 11:37:39,262 - INFO - machamp.modules.allennlp.slanted_triangular - 0: 6.25e-07
2025-07-10 11:37:39,262 - INFO - machamp.modules.allennlp.slanted_triangular - 1: 6.25e-07
2025-07-10 11:37:39,262 - INFO - machamp.modules.allennlp.slanted_triangular - 2: 2.3750000000000003e-07
2025-07-10 11:37:39,262 - INFO - machamp.modules.allennlp.slanted_triangular - Gradual unfreezing finished. Training all layers.
2025-07-10 11:37:39,262 - INFO - machamp.model.trainer - Epoch 60: evaluating on dev
2025-07-10 11:37:40,833 - INFO - machamp.model.callback - epoch       : 60/60
2025-07-10 11:37:40,833 - INFO - machamp.model.callback - best_epoch  : 59
2025-07-10 11:37:40,833 - INFO - machamp.model.callback - max_gpu_mem : 5.6741
2025-07-10 11:37:40,833 - INFO - machamp.model.callback - cur_ram     : 2.1411
2025-07-10 11:37:40,833 - INFO - machamp.model.callback - time_epoch  : 0:00:28
2025-07-10 11:37:40,833 - INFO - machamp.model.callback - time_total  : 0:29:31
2025-07-10 11:37:40,833 - INFO - machamp.model.callback - 
2025-07-10 11:37:40,834 - INFO - machamp.model.callback -                train_loss dev_loss train_scores dev_scores 
2025-07-10 11:37:40,834 - INFO - machamp.model.callback - Best (59)                                                  
2025-07-10 11:37:40,834 - INFO - machamp.model.callback - dependency_las     0.1509   2.2613       0.8962     0.6374 
2025-07-10 11:37:40,834 - INFO - machamp.model.callback - lemma_accuracy     0.3895   0.5621       0.8825     0.8618 
2025-07-10 11:37:40,834 - INFO - machamp.model.callback - morph_accuracy     0.6977   0.9341       0.8375     0.7689 
2025-07-10 11:37:40,834 - INFO - machamp.model.callback - upos_accuracy      0.0552   0.1384       0.9703     0.9141 
2025-07-10 11:37:40,834 - INFO - machamp.model.callback - xpos_accuracy      0.0942   0.1735       0.9519     0.8984 
2025-07-10 11:37:40,834 - INFO - machamp.model.callback - sum                1.3874   4.0695       4.5385     4.0805 
2025-07-10 11:37:40,834 - INFO - machamp.model.callback - Epoch 60                                                   
2025-07-10 11:37:40,834 - INFO - machamp.model.callback - dependency_las     0.1392   2.2589       0.9024     0.6370 
2025-07-10 11:37:40,834 - INFO - machamp.model.callback - lemma_accuracy     0.3889   0.5617       0.8814     0.8618 
2025-07-10 11:37:40,834 - INFO - machamp.model.callback - morph_accuracy     0.6951   0.9330       0.8384     0.7689 
2025-07-10 11:37:40,834 - INFO - machamp.model.callback - upos_accuracy      0.0552   0.1386       0.9693     0.9145 
2025-07-10 11:37:40,834 - INFO - machamp.model.callback - xpos_accuracy      0.0924   0.1736       0.9509     0.8984 
2025-07-10 11:37:40,834 - INFO - machamp.model.callback - sum                1.3708   4.0658       4.5425     4.0805 
2025-07-10 11:37:40,842 - INFO - machamp.model.callback - 
                Dev scores (y) over epochs (x)
┌────────────────────────────────────────────────────────────┐
│                     ▄▄▄▄▄▄▄▄▄▞▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄│
│                 ▄▛▀▄▄▄▄▀▀▀▀▀▀▀▄▄▄▄▄▄▄▄▄▄▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀│
│               ▄▗▄▀▀▄▄▄▛▀▀▀▀▀                       ▄▄▄▄▄▄▄▄│ 0.8
│            ▗▞▀▀▘▞▀▀          ▄▄▄▄▄▞▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀        │
│          ▗▄▘▘▄▀       ▄▄▞▀▀▀▀                              │
│   ▗▀▀▀▀▀▀▘▄▞▘    ▄▄▀▀▀▄▄▙▄▄▄▄▀▄▟▀▀▜▀▄▞▄▞▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀▀│ 0.6
│   ▌     ▞▘   ▗▄▀▀▀▀                                        │
│   ▌   ▗▞   ▄▄▀                                             │
│  ▗▘▛▀▛▘ ▗▞▄▀                                               │
│  ▐ ▗▞  ▄▘▞                                                 │ 0.4
│ ▗▐▀▘  ▞▗▀                                                  │
│▞▘▐▐ ▗▞▗▘                                                   │
│  ▌▐▄▄▄▘                                                    │
│  ▌▗▘                                                       │ 0.2
│▞▗▘▞                                                        │
│ ▞▗▘                                                        │
│▀▘▘                                                         │
└────────────────────────────────────────────────────────────┘
         10        20        30        40        50        60
   ██ dependency   ██ lemma   ██ morph   ██ upos   ██ xpos
2025-07-10 11:37:40,842 - INFO - machamp.model.callback - Best performance obtained in epoch 59 linking model model_59.pt as logs/ota_config/2025.07.10_11.07.44/model.pt.
2025-07-10 11:37:40,843 - INFO - machamp.model.trainer - 
2025-07-10 11:37:40,843 - ERROR - STDERR - /content/machamp/machamp/model/trainer.py:292: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.
2025-07-10 11:37:40,843 - ERROR - STDERR -   scalars[task] = torch.nn.functional.softmax(model.scalars[task].scalar_parameters.data).tolist()
2025-07-10 11:37:40,843 - INFO - machamp.model.trainer - Predicting on dev set
2025-07-10 11:37:43,147 - INFO - machamp.data.machamp_dataset - Reading /content/dev.conllu...
2025-07-10 11:37:43,161 - WARNING - machamp.data.machamp_vocabulary - Case=Acc|Number=Sing|Number[psor]=Sing|Person=3|Person[psor]=3|Polarity=Pos can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,202 - WARNING - machamp.data.machamp_vocabulary - Mood=Opt|Number=Plur|Person=1|Polarity=Neg|Tense=Pres|VerbForm=Fin can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,236 - WARNING - machamp.data.machamp_vocabulary - Aspect=Perf|Evident=Fh|Mood=Ind|Number=Plur|Person=3|Polarity=Pos|Tense=Past|VerbForm=Fin|Voice=Cau can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,264 - WARNING - machamp.data.machamp_vocabulary - Evident=Nfh|Number=Sing|Person=3|Polarity=Pos|Tense=Past|VerbForm=Fin can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,291 - WARNING - machamp.data.machamp_vocabulary - Aspect=Perf|Evident=Fh|Mood=Cnd|Number=Sing|Person=2|Polarity=Neg|Tense=Past|VerbForm=Fin can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,292 - WARNING - machamp.data.machamp_vocabulary - Mood=Ind|Number=Sing|Person=3|Tense=Pres|VerbForm=Fin can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,318 - WARNING - machamp.data.machamp_vocabulary - Aspect=Perf|Case=Nom|Number[psor]=Sing|Person[psor]=3|Polarity=Pos|Tense=Past|VerbForm=Part can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,372 - WARNING - machamp.data.machamp_vocabulary - Case=Nom|Number=Plur|Number[psor]=Plur|Person=3|Person[psor]=3|PronType=Prs|Reflex=Yes can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,388 - WARNING - machamp.data.machamp_vocabulary - Aspect=Imp|Polarity=Pos|Tense=Fut|VerbForm=Part can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,394 - WARNING - machamp.data.machamp_vocabulary - Case=Acc|Number=Sing|Number[psor]=Plur|Person=3|Person[psor]=3|Polarity=Pos|VerbForm=Vnoun can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,406 - WARNING - machamp.data.machamp_vocabulary - Case=Loc|Gender=Masc|Number=Sing|Person=3 can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,406 - WARNING - machamp.data.machamp_vocabulary - Aspect=Hab|Polarity=Neg|Tense=Pres|VerbForm=Part can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,413 - WARNING - machamp.data.machamp_vocabulary - Case=Nom|Number=Plur|Person=2 can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,427 - WARNING - machamp.data.machamp_vocabulary - Evident=Nfh|Number=Sing|Person=3|Polarity=Pos|Tense=Past|VerbForm=Fin can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,468 - WARNING - machamp.data.machamp_vocabulary - Case=Ins|Gender=Fem|Number=Sing|Number[psor]=Sing|Person=3|Person[psor]=3 can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,468 - WARNING - machamp.data.machamp_vocabulary - Aspect=Perf|Mood=Ind|Number=Plur|Person=3|Polarity=Pos|Tense=Past|VerbForm=Fin can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,472 - WARNING - machamp.data.machamp_vocabulary - Aspect=Perf|Mood=Ind|Number=Sing|Person=3|Polarity=Pos|Tense=Past|VerbForm=Fin|Voice=Pass can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,477 - WARNING - machamp.data.machamp_vocabulary - Aspect=Hab|Polarity=Neg|Tense=Pres|VerbForm=Part can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,501 - WARNING - machamp.data.machamp_vocabulary - Aspect=Perf|Case=Nom|Number[psor]=Sing|Person[psor]=3|Polarity=Pos|Tense=Pres|VerbForm=Vnoun can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,519 - WARNING - machamp.data.machamp_vocabulary - Case=Nom|Number=Plur|Person=3|Typo=Yes can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,558 - WARNING - machamp.data.machamp_vocabulary - Case=Acc|Gender=Masc|Number=Sing|Number[psor]=Sing|Person=3|Person[psor]=3|PronType=Ind can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,565 - WARNING - machamp.data.machamp_vocabulary - Without can not be found in namespace xpos. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,578 - WARNING - machamp.data.machamp_vocabulary - Aspect=Perf|Number[psor]=Sing|Person[psor]=3|Polarity=Pos|Tense=Past|VerbForm=Part|Voice=Pass can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,586 - WARNING - machamp.data.machamp_vocabulary - Case=Nom|Number=Sing|Number[psor]=Plur|Person=3|Person[psor]=3|Polarity=Pos|Tense=Pres|VerbForm=Part can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,599 - WARNING - machamp.data.machamp_vocabulary - Aspect=Hab|Polarity=Neg|VerbForm=Part can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,599 - WARNING - machamp.data.machamp_vocabulary - Aspect=Perf|Evident=Fh|Mood=Opt|Number=Sing|Person=3|Polarity=Pos|Tense=Past|VerbForm=Fin can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,621 - WARNING - machamp.data.machamp_vocabulary - Case=Loc|Number=Sing|Person=3|Polarity=Neg|VerbForm=Vnoun can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,626 - WARNING - machamp.data.machamp_vocabulary - Aspect=Hab|Mood=Cnd|Number=Sing|Person=3|Polarity=Neg|Tense=Pres|VerbForm=Fin can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,637 - WARNING - machamp.data.machamp_vocabulary - Case=Abl|Number=Sing|Person=3|Polarity=Pos|Tense=Pres|VerbForm=Part can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,649 - WARNING - machamp.data.machamp_vocabulary - Evident=Nfh|Number=Sing|Person=3|Polarity=Neg|Tense=Past|VerbForm=Fin can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,659 - WARNING - machamp.data.machamp_vocabulary - Case=Gen|Number=Sing|Number[psor]=Sing|Person=3|Person[psor]=3|PronType=Dem can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,665 - WARNING - machamp.data.machamp_vocabulary - Evident=Nfh|Number=Sing|Person=3|Polarity=Neg|Tense=Past|VerbForm=Part can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,668 - WARNING - machamp.data.machamp_vocabulary - Case=Abl|Number=Sing|Person=3|PronType=Ind can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,747 - WARNING - machamp.data.machamp_vocabulary - Case=Acc|Number=Sing|Number[psor]=Sing|Person=3|Person[psor]=3|Polarity=Pos|VerbForm=Vnoun|Voice=Pass can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,770 - WARNING - machamp.data.machamp_vocabulary - Aspect=Perf|Evident=Fh|Number=Sing|Person=3|Polarity=Pos|Tense=Past|VerbForm=Fin can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,790 - WARNING - machamp.data.machamp_vocabulary - Case=Nom|Gender=Fem|Number=Dual|Person=3 can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,790 - WARNING - machamp.data.machamp_vocabulary - Gender=Fem|Number=Dual can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,806 - WARNING - machamp.data.machamp_vocabulary - Aspect=Hab|Mood=Ind|Number=Plur|Person=3|Polarity=Pos|Tense=Pres|VerbForm=Fin|Voice=Pass can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,837 - WARNING - machamp.data.machamp_vocabulary - Case=Acc|Number=Sing|Number[psor]=Sing|Person=3|Person[psor]=2|Polarity=Pos|VerbForm=Vnoun can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,841 - WARNING - machamp.data.machamp_vocabulary - Evident=Nfh|Mood=Ind|Number=Plur|Person=3|Polarity=Pos|Tense=Past|VerbForm=Fin can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,905 - WARNING - machamp.data.machamp_vocabulary - Aspect=Hab|Mood=Cnd|Number=Sing|Person=1|Polarity=Neg|Tense=Pres|VerbForm=Fin can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,918 - WARNING - machamp.data.machamp_vocabulary - Aspect=Perf|Case=Dat|Number=Plur|Number[psor]=Plur|Person=3|Person[psor]=3|Polarity=Neg|Tense=Past|VerbForm=Part can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:43,944 - WARNING - machamp.data.machamp_vocabulary - Number=Sing|Person=2|Polarity=Pos|Tense=Pres can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:44,005 - WARNING - machamp.data.machamp_vocabulary - Aspect=Prosp|Mood=Ind|Polarity=Pos|Tense=Fut|VerbForm=Part can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:44,018 - WARNING - machamp.data.machamp_vocabulary - Case=Ins|Gender=Fem|Number=Sing|Number[psor]=Sing|Person=3|Person[psor]=3 can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:44,027 - WARNING - machamp.data.machamp_vocabulary - Aspect=Imp|Mood=Ind|Number=Sing|Person=3|Polarity=Pos|VerbForm=Fin can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:44,058 - WARNING - machamp.data.machamp_vocabulary - Case=Ins|Number=Sing|Person=3|Polarity=Neg|VerbForm=Vnoun can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:44,084 - WARNING - machamp.data.machamp_vocabulary - Case=Ins|Gender=Fem|Number=Sing|Number[psor]=Sing|Person=3|Person[psor]=1 can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:44,084 - WARNING - machamp.data.machamp_vocabulary - Case=Loc|Gender=Fem|Number=Sing|Number[psor]=Plur|Person=3|Person[psor]=1 can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:44,139 - WARNING - machamp.data.machamp_vocabulary - Evident=Nfh|Mood=Ind|Number=Sing|Person=3|Polarity=Pos|Tense=Past|VerbForm=Fin|Voice=Pass can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:44,181 - WARNING - machamp.data.machamp_vocabulary - Case=Ins|Number=Sing|Person=3|PronType=Dem can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:44,190 - WARNING - machamp.data.machamp_vocabulary - Aspect=Hab|Mood=Pot|Number=Plur|Person=2|Polarity=Pos|Tense=Pres|VerbForm=Fin can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:44,197 - WARNING - machamp.data.machamp_vocabulary - Case=Gen|Number=Plur|Number[psor]=Plur|Person=1|Person[psor]=1 can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:44,229 - WARNING - machamp.data.machamp_vocabulary - Case=Ins|Number=Sing|Person=3|Polarity=Pos|VerbForm=Vnoun|Voice=Pass can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:44,241 - WARNING - machamp.data.machamp_vocabulary - Aspect=Hab|Mood=Ind|Polarity=Pos|Tense=Pres|VerbForm=Part can not be found in namespace morph. This usually means you have unseen labels in your data during prediction. This will lead to incorrect macro-f1 scores.
2025-07-10 11:37:44,241 - INFO - machamp.readers.read_sequence - Stats UD-EWT (/content/dev.conllu):
2025-07-10 11:37:44,242 - INFO - machamp.readers.read_sequence - Lines:      291
2025-07-10 11:37:44,242 - INFO - machamp.readers.read_sequence - Words:      2,981
2025-07-10 11:37:44,242 - INFO - machamp.readers.read_sequence - Subwords:   9,043
2025-07-10 11:37:44,242 - INFO - machamp.readers.read_sequence - Unks:       130
2025-07-10 11:37:44,242 - INFO - machamp.readers.read_sequence - Pre-splits: 0
2025-07-10 11:37:44,242 - INFO - machamp.data.machamp_dataset - Done reading /content/dev.conllu (1.0s)

2025-07-10 11:37:46,668 - INFO - machamp.utils.myutils - dependency_las : 0.6374
2025-07-10 11:37:46,668 - INFO - machamp.utils.myutils - lemma_accuracy : 0.8618
2025-07-10 11:37:46,668 - INFO - machamp.utils.myutils - morph_accuracy : 0.7689
2025-07-10 11:37:46,668 - INFO - machamp.utils.myutils - upos_accuracy  : 0.9141
2025-07-10 11:37:46,668 - INFO - machamp.utils.myutils - xpos_accuracy  : 0.8984
2025-07-10 11:37:46,668 - INFO - machamp.utils.myutils - sum            : 4.0805
2025-07-10 11:37:46,668 - INFO - machamp.utils.myutils -