End of training

README.md

@@ -0,0 +1,117 @@
+---
+license: apache-2.0
+base_model: sentence-transformers/paraphrase-MiniLM-L3-v2
+tags:
+- generated_from_trainer
+datasets:
+- nyt_ingredients
+model-index:
+- name: nyt_ingredients-tagger-paraphrase-MiniLM-L3-v2
+  results: []
+---
+
+<!-- This model card has been generated automatically according to the information the Trainer had access to. You
+should probably proofread and complete it, then remove this comment. -->
+
+# nyt_ingredients-tagger-paraphrase-MiniLM-L3-v2
+
+This model is a fine-tuned version of [sentence-transformers/paraphrase-MiniLM-L3-v2](https://huggingface.co/sentence-transformers/paraphrase-MiniLM-L3-v2) on the nyt_ingredients dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.4826
+- Comment: {'precision': 0.641354292623942, 'recall': 0.7546955036994878, 'f1': 0.6934239769904563, 'number': 7028}
+- Name: {'precision': 0.7884535655058044, 'recall': 0.8198965294244449, 'f1': 0.8038676952340695, 'number': 9278}
+- Qty: {'precision': 0.9891548736939558, 'recall': 0.9878483687755911, 'f1': 0.9885011895321174, 'number': 7571}
+- Range End: {'precision': 0.813953488372093, 'recall': 0.7526881720430108, 'f1': 0.782122905027933, 'number': 93}
+- Unit: {'precision': 0.9268482490272374, 'recall': 0.9833223249669749, 'f1': 0.9542504607002644, 'number': 6056}
+- Overall Precision: 0.8257
+- Overall Recall: 0.8797
+- Overall F1: 0.8519
+- Overall Accuracy: 0.8289
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- learning_rate: 5e-05
+- train_batch_size: 32
+- eval_batch_size: 32
+- seed: 42
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: linear
+- num_epochs: 10
+
+### Training results
+
+| Training Loss | Epoch | Step  | Validation Loss | Comment                                                                                                   | Name                                                                                                      | Qty                                                                                                       | Range End                                                                                                | Unit                                                                                                      | Overall Precision | Overall Recall | Overall F1 | Overall Accuracy |
+|:-------------:|:-----:|:-----:|:---------------:|:---------------------------------------------------------------------------------------------------------:|:---------------------------------------------------------------------------------------------------------:|:---------------------------------------------------------------------------------------------------------:|:--------------------------------------------------------------------------------------------------------:|:---------------------------------------------------------------------------------------------------------:|:-----------------:|:--------------:|:----------:|:----------------:|
+| 0.7174        | 0.2   | 1000  | 0.6566          | {'precision': 0.5088729580444236, 'recall': 0.6345252051582649, 'f1': 0.5647948868453662, 'number': 6824} | {'precision': 0.7654784240150094, 'recall': 0.7879132114052028, 'f1': 0.7765338110165697, 'number': 8803} | {'precision': 0.9619699042407661, 'recall': 0.9810267857142857, 'f1': 0.9714048901782015, 'number': 7168} | {'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 82}                                               | {'precision': 0.9151633349585568, 'recall': 0.9784535186794092, 'f1': 0.945750755794424, 'number': 5755}  | 0.7711            | 0.8357         | 0.8021     | 0.7841           |
+| 0.6076        | 0.4   | 2000  | 0.5845          | {'precision': 0.5363414634146342, 'recall': 0.6444900351699883, 'f1': 0.5854632587859425, 'number': 6824} | {'precision': 0.7547189819724284, 'recall': 0.8084743837328183, 'f1': 0.7806724071738058, 'number': 8803} | {'precision': 0.974110480409802, 'recall': 0.9815848214285714, 'f1': 0.9778333680772705, 'number': 7168}  | {'precision': 0.6, 'recall': 0.14634146341463414, 'f1': 0.23529411764705882, 'number': 82}               | {'precision': 0.9114658925979681, 'recall': 0.9821025195482189, 'f1': 0.9454667112746737, 'number': 5755} | 0.7793            | 0.8457         | 0.8111     | 0.7988           |
+| 0.5777        | 0.59  | 3000  | 0.5503          | {'precision': 0.55049786628734, 'recall': 0.6805392731535757, 'f1': 0.6086500655307995, 'number': 6824}   | {'precision': 0.7717766278568349, 'recall': 0.8132454844939225, 'f1': 0.7919685823330935, 'number': 8803} | {'precision': 0.9715620277510647, 'recall': 0.9866071428571429, 'f1': 0.9790267875683533, 'number': 7168} | {'precision': 0.45714285714285713, 'recall': 0.5853658536585366, 'f1': 0.5133689839572192, 'number': 82} | {'precision': 0.9229885057471264, 'recall': 0.9767158992180712, 'f1': 0.9490924440692275, 'number': 5755} | 0.7870            | 0.8572         | 0.8206     | 0.8056           |
+| 0.553         | 0.79  | 4000  | 0.5343          | {'precision': 0.5713428537141486, 'recall': 0.6976846424384525, 'f1': 0.6282245827010623, 'number': 6824} | {'precision': 0.7823947512301804, 'recall': 0.812791093945246, 'f1': 0.7973033207042568, 'number': 8803}  | {'precision': 0.9756500206355757, 'recall': 0.9893973214285714, 'f1': 0.9824755835699939, 'number': 7168} | {'precision': 0.5465116279069767, 'recall': 0.573170731707317, 'f1': 0.5595238095238095, 'number': 82}   | {'precision': 0.9150137074665377, 'recall': 0.9859252823631625, 'f1': 0.9491468718634996, 'number': 5755} | 0.7968            | 0.8637         | 0.8289     | 0.8122           |
+| 0.5407        | 0.99  | 5000  | 0.5183          | {'precision': 0.5958625785001848, 'recall': 0.7091148886283705, 'f1': 0.6475744396119103, 'number': 6824} | {'precision': 0.787269129287599, 'recall': 0.8134726797682609, 'f1': 0.8001564333202973, 'number': 8803}  | {'precision': 0.9823488533703961, 'recall': 0.9860491071428571, 'f1': 0.9841955023323818, 'number': 7168} | {'precision': 0.5294117647058824, 'recall': 0.7682926829268293, 'f1': 0.626865671641791, 'number': 82}   | {'precision': 0.9231271421576628, 'recall': 0.9827975673327541, 'f1': 0.9520282780676653, 'number': 5755} | 0.8085            | 0.8657         | 0.8361     | 0.8186           |
+| 0.5109        | 1.19  | 6000  | 0.5210          | {'precision': 0.611145206824598, 'recall': 0.729630715123095, 'f1': 0.6651526284149355, 'number': 6824}   | {'precision': 0.788943623426382, 'recall': 0.8186981710780415, 'f1': 0.803545545768759, 'number': 8803}   | {'precision': 0.9822370247016375, 'recall': 0.9874441964285714, 'f1': 0.9848337275636566, 'number': 7168} | {'precision': 0.5614035087719298, 'recall': 0.7804878048780488, 'f1': 0.653061224489796, 'number': 82}   | {'precision': 0.924762839385018, 'recall': 0.9824500434404866, 'f1': 0.9527340129749768, 'number': 5755}  | 0.8133            | 0.8725         | 0.8419     | 0.8201           |
+| 0.5125        | 1.39  | 7000  | 0.5072          | {'precision': 0.607906114885732, 'recall': 0.7211313012895663, 'f1': 0.6596956900596554, 'number': 6824}  | {'precision': 0.7913407513495648, 'recall': 0.8159718277859821, 'f1': 0.8034675615212529, 'number': 8803} | {'precision': 0.983300862788756, 'recall': 0.9857700892857143, 'f1': 0.9845339278249966, 'number': 7168}  | {'precision': 0.5714285714285714, 'recall': 0.7317073170731707, 'f1': 0.641711229946524, 'number': 82}   | {'precision': 0.9279947273026857, 'recall': 0.978627280625543, 'f1': 0.952638700947226, 'number': 5755}   | 0.8143            | 0.8683         | 0.8404     | 0.8215           |
+| 0.4967        | 1.58  | 8000  | 0.5088          | {'precision': 0.6126803996546195, 'recall': 0.7278722157092614, 'f1': 0.6653271716562856, 'number': 6824} | {'precision': 0.7949600355239786, 'recall': 0.8134726797682609, 'f1': 0.8041098197742967, 'number': 8803} | {'precision': 0.9813148788927336, 'recall': 0.9891183035714286, 'f1': 0.9852011394427846, 'number': 7168} | {'precision': 0.503448275862069, 'recall': 0.8902439024390244, 'f1': 0.6431718061674009, 'number': 82}   | {'precision': 0.9266295447101212, 'recall': 0.9831450912250217, 'f1': 0.9540510918135063, 'number': 5755} | 0.8156            | 0.8714         | 0.8425     | 0.8218           |
+| 0.5121        | 1.78  | 9000  | 0.5022          | {'precision': 0.6241911398705824, 'recall': 0.735052754982415, 'f1': 0.6751009421265142, 'number': 6824}  | {'precision': 0.7883291443558081, 'recall': 0.8194933545382256, 'f1': 0.803609223571349, 'number': 8803}  | {'precision': 0.9850642099385818, 'recall': 0.9845145089285714, 'f1': 0.9847892827239743, 'number': 7168} | {'precision': 0.5238095238095238, 'recall': 0.8048780487804879, 'f1': 0.6346153846153846, 'number': 82}  | {'precision': 0.9253999347045381, 'recall': 0.9850564726324935, 'f1': 0.9542967763656258, 'number': 5755} | 0.8176            | 0.8739         | 0.8448     | 0.8248           |
+| 0.5008        | 1.98  | 10000 | 0.4998          | {'precision': 0.6337506387327542, 'recall': 0.7269929660023446, 'f1': 0.6771771771771772, 'number': 6824} | {'precision': 0.7910726036411494, 'recall': 0.8193797569010565, 'f1': 0.8049774008146867, 'number': 8803} | {'precision': 0.9698246567894522, 'recall': 0.9953962053571429, 'f1': 0.9824440619621343, 'number': 7168} | {'precision': 0.47468354430379744, 'recall': 0.9146341463414634, 'f1': 0.625, 'number': 82}              | {'precision': 0.9263984298331698, 'recall': 0.9841876629018245, 'f1': 0.9544190749010025, 'number': 5755} | 0.8192            | 0.8748         | 0.8461     | 0.8250           |
+| 0.4863        | 2.18  | 11000 | 0.4954          | {'precision': 0.6343217665615142, 'recall': 0.736664712778429, 'f1': 0.6816733337853415, 'number': 6824}  | {'precision': 0.7913881888901063, 'recall': 0.820515733272748, 'f1': 0.8056887897378694, 'number': 8803}  | {'precision': 0.9850787895690978, 'recall': 0.9854910714285714, 'f1': 0.9852848873701096, 'number': 7168} | {'precision': 0.6122448979591837, 'recall': 0.7317073170731707, 'f1': 0.6666666666666666, 'number': 82}  | {'precision': 0.9250367466927977, 'recall': 0.9841876629018245, 'f1': 0.9536959084020878, 'number': 5755} | 0.8224            | 0.8745         | 0.8477     | 0.8273           |
+| 0.4893        | 2.38  | 12000 | 0.4947          | {'precision': 0.6325136612021858, 'recall': 0.7463364595545134, 'f1': 0.6847270771712826, 'number': 6824} | {'precision': 0.7868191721132898, 'recall': 0.820515733272748, 'f1': 0.8033142412278262, 'number': 8803}  | {'precision': 0.9809207797594359, 'recall': 0.9898158482142857, 'f1': 0.9853482397055761, 'number': 7168} | {'precision': 0.5068493150684932, 'recall': 0.9024390243902439, 'f1': 0.6491228070175438, 'number': 82}  | {'precision': 0.9245928338762215, 'recall': 0.9864465682015638, 'f1': 0.9545187053383775, 'number': 5755} | 0.8182            | 0.8788         | 0.8474     | 0.8257           |
+| 0.489         | 2.57  | 13000 | 0.4869          | {'precision': 0.635054931178179, 'recall': 0.736957796014068, 'f1': 0.6822220714915552, 'number': 6824}   | {'precision': 0.7892552959161389, 'recall': 0.8210837214585937, 'f1': 0.8048549635320973, 'number': 8803} | {'precision': 0.9829498197948433, 'recall': 0.9892578125, 'f1': 0.9860937282714505, 'number': 7168}       | {'precision': 0.5298507462686567, 'recall': 0.8658536585365854, 'f1': 0.6574074074074074, 'number': 82}  | {'precision': 0.9235896602178507, 'recall': 0.9871416159860991, 'f1': 0.9543087518898035, 'number': 5755} | 0.8209            | 0.8766         | 0.8479     | 0.8272           |
+| 0.4691        | 2.77  | 14000 | 0.4905          | {'precision': 0.6450718427022939, 'recall': 0.75, 'f1': 0.6935899173329719, 'number': 6824}               | {'precision': 0.7925502692011867, 'recall': 0.8193797569010565, 'f1': 0.8057417336907953, 'number': 8803} | {'precision': 0.9818584683561834, 'recall': 0.9891183035714286, 'f1': 0.9854750156369451, 'number': 7168} | {'precision': 0.5481481481481482, 'recall': 0.9024390243902439, 'f1': 0.6820276497695853, 'number': 82}  | {'precision': 0.9296515450361604, 'recall': 0.9827975673327541, 'f1': 0.9554861052453754, 'number': 5755} | 0.8253            | 0.8784         | 0.8510     | 0.8283           |
+| 0.4781        | 2.97  | 15000 | 0.4878          | {'precision': 0.6444919517102615, 'recall': 0.7510257913247362, 'f1': 0.6936924742826204, 'number': 6824} | {'precision': 0.7942146942366916, 'recall': 0.8202885379984096, 'f1': 0.8070410729253981, 'number': 8803} | {'precision': 0.98375, 'recall': 0.9881417410714286, 'f1': 0.9859409799554566, 'number': 7168}            | {'precision': 0.5433070866141733, 'recall': 0.8414634146341463, 'f1': 0.6602870813397129, 'number': 82}  | {'precision': 0.9268651832460733, 'recall': 0.9843614248479583, 'f1': 0.9547484621218505, 'number': 5755} | 0.8255            | 0.8788         | 0.8513     | 0.8293           |
+| 0.4599        | 3.17  | 16000 | 0.4852          | {'precision': 0.6450299096347206, 'recall': 0.742672919109027, 'f1': 0.6904161841836387, 'number': 6824}  | {'precision': 0.7944048272078991, 'recall': 0.8225604907417926, 'f1': 0.8082375265096551, 'number': 8803} | {'precision': 0.983072013320383, 'recall': 0.9884207589285714, 'f1': 0.9857391304347826, 'number': 7168}  | {'precision': 0.5289855072463768, 'recall': 0.8902439024390244, 'f1': 0.6636363636363637, 'number': 82}  | {'precision': 0.9331015068719987, 'recall': 0.9791485664639444, 'f1': 0.9555706291334577, 'number': 5755} | 0.8269            | 0.8767         | 0.8511     | 0.8306           |
+| 0.4636        | 3.37  | 17000 | 0.4889          | {'precision': 0.6520854526958291, 'recall': 0.7514654161781946, 'f1': 0.6982570806100218, 'number': 6824} | {'precision': 0.7879774305555556, 'recall': 0.8249460411223447, 'f1': 0.8060380709251347, 'number': 8803} | {'precision': 0.9818785447503112, 'recall': 0.990234375, 'f1': 0.9860387580745988, 'number': 7168}        | {'precision': 0.46875, 'recall': 0.9146341463414634, 'f1': 0.6198347107438017, 'number': 82}             | {'precision': 0.9251711770459733, 'recall': 0.9860990443092963, 'f1': 0.9546639751030365, 'number': 5755} | 0.8247            | 0.8815         | 0.8521     | 0.8303           |
+| 0.4688        | 3.56  | 18000 | 0.4901          | {'precision': 0.6491074772188241, 'recall': 0.7620164126611958, 'f1': 0.7010448264239973, 'number': 6824} | {'precision': 0.7955320787938813, 'recall': 0.8211973190957628, 'f1': 0.8081609837898266, 'number': 8803} | {'precision': 0.9829474559822543, 'recall': 0.9891183035714286, 'f1': 0.9860232250886588, 'number': 7168} | {'precision': 0.5526315789473685, 'recall': 0.7682926829268293, 'f1': 0.6428571428571429, 'number': 82}  | {'precision': 0.9250325945241199, 'recall': 0.98627280625543, 'f1': 0.9546716003700276, 'number': 5755}   | 0.8265            | 0.8822         | 0.8534     | 0.8319           |
+| 0.4584        | 3.76  | 19000 | 0.4824          | {'precision': 0.65155342258146, 'recall': 0.7560082063305978, 'f1': 0.6999050332383666, 'number': 6824}   | {'precision': 0.7932665859830564, 'recall': 0.819038963989549, 'f1': 0.8059467918622849, 'number': 8803}  | {'precision': 0.981740213030848, 'recall': 0.9900948660714286, 'f1': 0.9858998402444954, 'number': 7168}  | {'precision': 0.5289855072463768, 'recall': 0.8902439024390244, 'f1': 0.6636363636363637, 'number': 82}  | {'precision': 0.9315113598946329, 'recall': 0.9831450912250217, 'f1': 0.956632006086736, 'number': 5755}  | 0.8275            | 0.8800         | 0.8530     | 0.8307           |
+| 0.4701        | 3.96  | 20000 | 0.4806          | {'precision': 0.6606470053267507, 'recall': 0.7451641266119577, 'f1': 0.7003649886371462, 'number': 6824} | {'precision': 0.7948661693725319, 'recall': 0.8231284789276383, 'f1': 0.8087504883084994, 'number': 8803} | {'precision': 0.9844357976653697, 'recall': 0.98828125, 'f1': 0.98635477582846, 'number': 7168}           | {'precision': 0.5384615384615384, 'recall': 0.8536585365853658, 'f1': 0.660377358490566, 'number': 82}   | {'precision': 0.9314549349151425, 'recall': 0.9822762814943528, 'f1': 0.956190798376184, 'number': 5755}  | 0.8322            | 0.8780         | 0.8545     | 0.8329           |
+| 0.4441        | 4.16  | 21000 | 0.4863          | {'precision': 0.6520674996779595, 'recall': 0.7417936694021102, 'f1': 0.6940426407074792, 'number': 6824} | {'precision': 0.7900534409423056, 'recall': 0.8229012836533001, 'f1': 0.8061428889383486, 'number': 8803} | {'precision': 0.9831851028349082, 'recall': 0.9870256696428571, 'f1': 0.9851016429963798, 'number': 7168} | {'precision': 0.5882352941176471, 'recall': 0.7317073170731707, 'f1': 0.6521739130434783, 'number': 82}  | {'precision': 0.9332010582010583, 'recall': 0.9807124239791486, 'f1': 0.9563670253325427, 'number': 5755} | 0.8285            | 0.8761         | 0.8516     | 0.8310           |
+| 0.4375        | 4.36  | 22000 | 0.4853          | {'precision': 0.6539783577339274, 'recall': 0.7527842907385698, 'f1': 0.699911438108863, 'number': 6824}  | {'precision': 0.7924075488164066, 'recall': 0.8251732363966829, 'f1': 0.8084585420144685, 'number': 8803} | {'precision': 0.9818735298187353, 'recall': 0.9899553571428571, 'f1': 0.985897881208753, 'number': 7168}  | {'precision': 0.5103448275862069, 'recall': 0.9024390243902439, 'f1': 0.6519823788546256, 'number': 82}  | {'precision': 0.9296490652673007, 'recall': 0.9850564726324935, 'f1': 0.9565510841137266, 'number': 5755} | 0.8278            | 0.8815         | 0.8538     | 0.8326           |
+| 0.4566        | 4.55  | 23000 | 0.4807          | {'precision': 0.6580237358101135, 'recall': 0.7475087924970691, 'f1': 0.6999176728869374, 'number': 6824} | {'precision': 0.7963125548726954, 'recall': 0.8242644552993298, 'f1': 0.8100474462740721, 'number': 8803} | {'precision': 0.9836270292770917, 'recall': 0.9889787946428571, 'f1': 0.9862956521739129, 'number': 7168} | {'precision': 0.6090909090909091, 'recall': 0.8170731707317073, 'f1': 0.6979166666666666, 'number': 82}  | {'precision': 0.9343936381709742, 'recall': 0.9800173761946134, 'f1': 0.9566618607412434, 'number': 5755} | 0.8324            | 0.8785         | 0.8548     | 0.8325           |
+| 0.4546        | 4.75  | 24000 | 0.4797          | {'precision': 0.6631361405321622, 'recall': 0.7523446658851114, 'f1': 0.7049292873815736, 'number': 6824} | {'precision': 0.7910382513661203, 'recall': 0.8222196978302851, 'f1': 0.8063276332646355, 'number': 8803} | {'precision': 0.985369931726348, 'recall': 0.9866071428571429, 'f1': 0.9859881491808993, 'number': 7168}  | {'precision': 0.651685393258427, 'recall': 0.7073170731707317, 'f1': 0.6783625730994152, 'number': 82}   | {'precision': 0.9321222130470685, 'recall': 0.9807124239791486, 'f1': 0.9558001693480102, 'number': 5755} | 0.8323            | 0.8782         | 0.8547     | 0.8324           |
+| 0.4462        | 4.95  | 25000 | 0.4855          | {'precision': 0.6583979328165375, 'recall': 0.7467760844079718, 'f1': 0.6998077451249656, 'number': 6824} | {'precision': 0.7887430864331417, 'recall': 0.8261956151312053, 'f1': 0.8070350643586329, 'number': 8803} | {'precision': 0.9822788315104527, 'recall': 0.9898158482142857, 'f1': 0.9860329372524493, 'number': 7168} | {'precision': 0.5390625, 'recall': 0.8414634146341463, 'f1': 0.6571428571428573, 'number': 82}           | {'precision': 0.9290873276428102, 'recall': 0.9834926151172894, 'f1': 0.9555161644298135, 'number': 5755} | 0.8286            | 0.8799         | 0.8535     | 0.8325           |
+| 0.433         | 5.15  | 26000 | 0.4934          | {'precision': 0.656934306569343, 'recall': 0.7517584994138335, 'f1': 0.7011549238023645, 'number': 6824}  | {'precision': 0.7907638432917488, 'recall': 0.8208565261842554, 'f1': 0.805529234713784, 'number': 8803}  | {'precision': 0.9852367688022284, 'recall': 0.9868861607142857, 'f1': 0.9860607750209088, 'number': 7168} | {'precision': 0.5172413793103449, 'recall': 0.9146341463414634, 'f1': 0.6607929515418502, 'number': 82}  | {'precision': 0.935024034477043, 'recall': 0.9801911381407472, 'f1': 0.9570749915167968, 'number': 5755}  | 0.8298            | 0.8782         | 0.8533     | 0.8298           |
+| 0.442         | 5.35  | 27000 | 0.4850          | {'precision': 0.6560364464692483, 'recall': 0.7596717467760844, 'f1': 0.7040608447643624, 'number': 6824} | {'precision': 0.7970520294797052, 'recall': 0.8231284789276383, 'f1': 0.8098804068402817, 'number': 8803} | {'precision': 0.9842902822188239, 'recall': 0.9877232142857143, 'f1': 0.9860037601838312, 'number': 7168} | {'precision': 0.5407407407407407, 'recall': 0.8902439024390244, 'f1': 0.6728110599078342, 'number': 82}  | {'precision': 0.9294735115630638, 'recall': 0.9847089487402259, 'f1': 0.956294296321296, 'number': 5755}  | 0.8301            | 0.8819         | 0.8552     | 0.8333           |
+| 0.4348        | 5.54  | 28000 | 0.4909          | {'precision': 0.6646636185499674, 'recall': 0.7456037514654161, 'f1': 0.7028109676082602, 'number': 6824} | {'precision': 0.7959763831183031, 'recall': 0.8269907985913894, 'f1': 0.8111872527717422, 'number': 8803} | {'precision': 0.982281284606866, 'recall': 0.9899553571428571, 'f1': 0.9861033907726514, 'number': 7168}  | {'precision': 0.5867768595041323, 'recall': 0.8658536585365854, 'f1': 0.6995073891625616, 'number': 82}  | {'precision': 0.9282555282555283, 'recall': 0.9847089487402259, 'f1': 0.9556492411467116, 'number': 5755} | 0.8331            | 0.8802         | 0.8560     | 0.8320           |
+| 0.4324        | 5.74  | 29000 | 0.4822          | {'precision': 0.6659320715582059, 'recall': 0.7527842907385698, 'f1': 0.7066996835878389, 'number': 6824} | {'precision': 0.7957977675640184, 'recall': 0.8260820174940361, 'f1': 0.8106571540047935, 'number': 8803} | {'precision': 0.9828278631768453, 'recall': 0.9900948660714286, 'f1': 0.9864479810966712, 'number': 7168} | {'precision': 0.5982905982905983, 'recall': 0.8536585365853658, 'f1': 0.7035175879396985, 'number': 82}  | {'precision': 0.9346850589995014, 'recall': 0.9772371850564726, 'f1': 0.9554875976894326, 'number': 5755} | 0.8342            | 0.8801         | 0.8566     | 0.8341           |
+| 0.4375        | 5.94  | 30000 | 0.4843          | {'precision': 0.6641568220173958, 'recall': 0.7497069167643611, 'f1': 0.70434363598816, 'number': 6824}   | {'precision': 0.7948914711686034, 'recall': 0.823696467113484, 'f1': 0.8090376569037656, 'number': 8803}  | {'precision': 0.9834836918806384, 'recall': 0.9885602678571429, 'f1': 0.9860154456272177, 'number': 7168} | {'precision': 0.5352112676056338, 'recall': 0.926829268292683, 'f1': 0.6785714285714285, 'number': 82}   | {'precision': 0.9328703703703703, 'recall': 0.9803649000868809, 'f1': 0.9560281284419215, 'number': 5755} | 0.8329            | 0.8791         | 0.8554     | 0.8328           |
+| 0.4226        | 6.14  | 31000 | 0.4878          | {'precision': 0.6603463531791177, 'recall': 0.7655334114888629, 'f1': 0.7090600610790635, 'number': 6824} | {'precision': 0.7907611663208475, 'recall': 0.8225604907417926, 'f1': 0.806347438752784, 'number': 8803}  | {'precision': 0.9829356270810211, 'recall': 0.9884207589285714, 'f1': 0.9856705620478576, 'number': 7168} | {'precision': 0.5241379310344828, 'recall': 0.926829268292683, 'f1': 0.669603524229075, 'number': 82}    | {'precision': 0.9298015417418403, 'recall': 0.9850564726324935, 'f1': 0.9566317921025986, 'number': 5755} | 0.8289            | 0.8835         | 0.8553     | 0.8336           |
+| 0.4305        | 6.34  | 32000 | 0.4837          | {'precision': 0.6565505182964906, 'recall': 0.770369284876905, 'f1': 0.7089205043489988, 'number': 6824}  | {'precision': 0.8005732554293904, 'recall': 0.8249460411223447, 'f1': 0.8125769273805529, 'number': 8803} | {'precision': 0.985369931726348, 'recall': 0.9866071428571429, 'f1': 0.9859881491808993, 'number': 7168}  | {'precision': 0.5528455284552846, 'recall': 0.8292682926829268, 'f1': 0.6634146341463415, 'number': 82}  | {'precision': 0.9309437684972048, 'recall': 0.9838401390095569, 'f1': 0.956661316211878, 'number': 5755}  | 0.8313            | 0.8844         | 0.8570     | 0.8342           |
+| 0.4142        | 6.53  | 33000 | 0.4862          | {'precision': 0.6611656843483955, 'recall': 0.7397420867526378, 'f1': 0.6982502247734974, 'number': 6824} | {'precision': 0.7957059918939643, 'recall': 0.8251732363966829, 'f1': 0.8101717599821548, 'number': 8803} | {'precision': 0.9761611179613646, 'recall': 0.9940011160714286, 'f1': 0.9850003456141564, 'number': 7168} | {'precision': 0.5323741007194245, 'recall': 0.9024390243902439, 'f1': 0.669683257918552, 'number': 82}   | {'precision': 0.9326843755155915, 'recall': 0.9822762814943528, 'f1': 0.9568381855111713, 'number': 5755} | 0.8315            | 0.8789         | 0.8545     | 0.8332           |
+| 0.4327        | 6.73  | 34000 | 0.4835          | {'precision': 0.6665352260778128, 'recall': 0.7431125439624854, 'f1': 0.7027439024390244, 'number': 6824} | {'precision': 0.7882020026121027, 'recall': 0.8226740883789617, 'f1': 0.8050692012673003, 'number': 8803} | {'precision': 0.9820317899101589, 'recall': 0.9912109375, 'f1': 0.9866000138859959, 'number': 7168}       | {'precision': 0.5833333333333334, 'recall': 0.8536585365853658, 'f1': 0.6930693069306931, 'number': 82}  | {'precision': 0.9306035191580332, 'recall': 0.9833188531711555, 'f1': 0.9562352145995269, 'number': 5755} | 0.8318            | 0.8783         | 0.8544     | 0.8330           |
+| 0.4299        | 6.93  | 35000 | 0.4833          | {'precision': 0.660639246147969, 'recall': 0.7602579132473622, 'f1': 0.706956462492335, 'number': 6824}   | {'precision': 0.7949644225506295, 'recall': 0.8249460411223447, 'f1': 0.8096777790166128, 'number': 8803} | {'precision': 0.9836133870295792, 'recall': 0.9881417410714286, 'f1': 0.9858723641171967, 'number': 7168} | {'precision': 0.5401459854014599, 'recall': 0.9024390243902439, 'f1': 0.6757990867579909, 'number': 82}  | {'precision': 0.9348582794629537, 'recall': 0.9800173761946134, 'f1': 0.9569053274516457, 'number': 5755} | 0.8316            | 0.8818         | 0.8560     | 0.8333           |
+| 0.4165        | 7.13  | 36000 | 0.4871          | {'precision': 0.6667525441195414, 'recall': 0.7584994138335287, 'f1': 0.7096729965037363, 'number': 6824} | {'precision': 0.7905082536924414, 'recall': 0.8268772009542201, 'f1': 0.8082838265504414, 'number': 8803} | {'precision': 0.9828278631768453, 'recall': 0.9900948660714286, 'f1': 0.9864479810966712, 'number': 7168} | {'precision': 0.5303030303030303, 'recall': 0.8536585365853658, 'f1': 0.6542056074766355, 'number': 82}  | {'precision': 0.930909689093601, 'recall': 0.9833188531711555, 'f1': 0.9563968227142133, 'number': 5755}  | 0.8315            | 0.8830         | 0.8565     | 0.8348           |
+| 0.4213        | 7.33  | 37000 | 0.4882          | {'precision': 0.667705088265836, 'recall': 0.753810082063306, 'f1': 0.7081497797356828, 'number': 6824}   | {'precision': 0.7934001306904814, 'recall': 0.827558786777235, 'f1': 0.8101195440644984, 'number': 8803}  | {'precision': 0.983754512635379, 'recall': 0.9884207589285714, 'f1': 0.9860821155184412, 'number': 7168}  | {'precision': 0.5333333333333333, 'recall': 0.8780487804878049, 'f1': 0.663594470046083, 'number': 82}   | {'precision': 0.9285831285831286, 'recall': 0.9850564726324935, 'f1': 0.9559865092748736, 'number': 5755} | 0.8327            | 0.8821         | 0.8567     | 0.8333           |
+| 0.4122        | 7.52  | 38000 | 0.4914          | {'precision': 0.660262119862578, 'recall': 0.7604044548651817, 'f1': 0.7068037866920929, 'number': 6824}  | {'precision': 0.7921752397558849, 'recall': 0.8257412245825286, 'f1': 0.8086100450525612, 'number': 8803} | {'precision': 0.9839090026355944, 'recall': 0.9895368303571429, 'f1': 0.9867148918411351, 'number': 7168} | {'precision': 0.5714285714285714, 'recall': 0.8292682926829268, 'f1': 0.6766169154228855, 'number': 82}  | {'precision': 0.9313693219223174, 'recall': 0.9833188531711555, 'f1': 0.9566393373341222, 'number': 5755} | 0.8304            | 0.8829         | 0.8559     | 0.8337           |
+| 0.4034        | 7.72  | 39000 | 0.4884          | {'precision': 0.6667092815136794, 'recall': 0.7642145369284877, 'f1': 0.7121398334016114, 'number': 6824} | {'precision': 0.7976099111939481, 'recall': 0.8264228104055436, 'f1': 0.8117607676857844, 'number': 8803} | {'precision': 0.984309913912802, 'recall': 0.9889787946428571, 'f1': 0.9866388308977034, 'number': 7168}  | {'precision': 0.568, 'recall': 0.8658536585365854, 'f1': 0.6859903381642511, 'number': 82}               | {'precision': 0.9304733727810651, 'recall': 0.9836663770634231, 'f1': 0.9563307711799983, 'number': 5755} | 0.8339            | 0.8840         | 0.8582     | 0.8344           |
+| 0.4209        | 7.92  | 40000 | 0.4910          | {'precision': 0.6634454322896534, 'recall': 0.7545427901524033, 'f1': 0.706067877956805, 'number': 6824}  | {'precision': 0.7984163642362256, 'recall': 0.8247188458480064, 'f1': 0.81135449262405, 'number': 8803}   | {'precision': 0.9832317073170732, 'recall': 0.9898158482142857, 'f1': 0.9865127919911012, 'number': 7168} | {'precision': 0.5413533834586466, 'recall': 0.8780487804878049, 'f1': 0.6697674418604651, 'number': 82}  | {'precision': 0.9315789473684211, 'recall': 0.9841876629018245, 'f1': 0.9571609632446134, 'number': 5755} | 0.8335            | 0.8815         | 0.8568     | 0.8327           |
+| 0.4078        | 8.12  | 41000 | 0.4923          | {'precision': 0.6621812123542228, 'recall': 0.7571805392731535, 'f1': 0.7065016749846176, 'number': 6824} | {'precision': 0.7941208611080757, 'recall': 0.8255140293081904, 'f1': 0.8095132004010249, 'number': 8803} | {'precision': 0.9834882752879145, 'recall': 0.9888392857142857, 'f1': 0.9861565217391305, 'number': 7168} | {'precision': 0.5581395348837209, 'recall': 0.8780487804878049, 'f1': 0.6824644549763034, 'number': 82}  | {'precision': 0.9316534914361001, 'recall': 0.982971329278888, 'f1': 0.9566246723598546, 'number': 5755}  | 0.8317            | 0.8819         | 0.8561     | 0.8335           |
+| 0.4098        | 8.32  | 42000 | 0.4903          | {'precision': 0.6613067891781521, 'recall': 0.7593786635404455, 'f1': 0.7069577080491133, 'number': 6824} | {'precision': 0.7961356899769458, 'recall': 0.8238100647506532, 'f1': 0.8097364895042429, 'number': 8803} | {'precision': 0.9838888888888889, 'recall': 0.98828125, 'f1': 0.9860801781737194, 'number': 7168}         | {'precision': 0.5669291338582677, 'recall': 0.8780487804878049, 'f1': 0.6889952153110047, 'number': 82}  | {'precision': 0.9302478253733792, 'recall': 0.9848827106863597, 'f1': 0.9567859554355165, 'number': 5755} | 0.8318            | 0.8822         | 0.8562     | 0.8344           |
+| 0.4009        | 8.51  | 43000 | 0.4936          | {'precision': 0.6647553516819572, 'recall': 0.7645076201641267, 'f1': 0.7111504907306434, 'number': 6824} | {'precision': 0.7969554265688314, 'recall': 0.8266500056798819, 'f1': 0.8115311698449873, 'number': 8803} | {'precision': 0.9840366463076069, 'recall': 0.9889787946428571, 'f1': 0.9865015307542444, 'number': 7168} | {'precision': 0.5689655172413793, 'recall': 0.8048780487804879, 'f1': 0.6666666666666667, 'number': 82}  | {'precision': 0.9306605323693723, 'recall': 0.9841876629018245, 'f1': 0.9566759564225994, 'number': 5755} | 0.8331            | 0.8841         | 0.8578     | 0.8348           |
+| 0.41          | 8.71  | 44000 | 0.4894          | {'precision': 0.6659480025364616, 'recall': 0.7694900351699883, 'f1': 0.7139846352573254, 'number': 6824} | {'precision': 0.8019158775600088, 'recall': 0.8273315915028967, 'f1': 0.8144254962258876, 'number': 8803} | {'precision': 0.9834974344751075, 'recall': 0.9893973214285714, 'f1': 0.9864385562278323, 'number': 7168} | {'precision': 0.5407407407407407, 'recall': 0.8902439024390244, 'f1': 0.6728110599078342, 'number': 82}  | {'precision': 0.9306946953522746, 'recall': 0.9847089487402259, 'f1': 0.9569402228976698, 'number': 5755} | 0.8344            | 0.8859         | 0.8594     | 0.8359           |
+| 0.4177        | 8.91  | 45000 | 0.4894          | {'precision': 0.6695323694151362, 'recall': 0.7532239155920282, 'f1': 0.7089166264395559, 'number': 6824} | {'precision': 0.7959273045763083, 'recall': 0.8258548222196979, 'f1': 0.810614930032893, 'number': 8803}  | {'precision': 0.9836270292770917, 'recall': 0.9889787946428571, 'f1': 0.9862956521739129, 'number': 7168} | {'precision': 0.5772357723577236, 'recall': 0.8658536585365854, 'f1': 0.6926829268292684, 'number': 82}  | {'precision': 0.9314370272936534, 'recall': 0.9843614248479583, 'f1': 0.9571682014023823, 'number': 5755} | 0.8350            | 0.8814         | 0.8575     | 0.8339           |
+| 0.399         | 9.11  | 46000 | 0.4909          | {'precision': 0.6659901052898642, 'recall': 0.7693434935521688, 'f1': 0.7139457401237506, 'number': 6824} | {'precision': 0.7953452797202797, 'recall': 0.8268772009542201, 'f1': 0.8108047897521581, 'number': 8803} | {'precision': 0.9826725810923205, 'recall': 0.9889787946428571, 'f1': 0.9858156028368794, 'number': 7168} | {'precision': 0.5488721804511278, 'recall': 0.8902439024390244, 'f1': 0.6790697674418604, 'number': 82}  | {'precision': 0.9298820445609436, 'recall': 0.98627280625543, 'f1': 0.9572476600050595, 'number': 5755}   | 0.8321            | 0.8860         | 0.8582     | 0.8356           |
+| 0.4066        | 9.31  | 47000 | 0.4892          | {'precision': 0.6655544147843943, 'recall': 0.7599648300117233, 'f1': 0.7096332785987959, 'number': 6824} | {'precision': 0.7964563053702286, 'recall': 0.8272179938657276, 'f1': 0.8115457483561797, 'number': 8803} | {'precision': 0.9824075356697604, 'recall': 0.9893973214285714, 'f1': 0.9858900396191006, 'number': 7168} | {'precision': 0.5538461538461539, 'recall': 0.8780487804878049, 'f1': 0.6792452830188679, 'number': 82}  | {'precision': 0.9298475659727914, 'recall': 0.9857515204170286, 'f1': 0.9569838056680162, 'number': 5755} | 0.8328            | 0.8838         | 0.8575     | 0.8340           |
+| 0.397         | 9.5   | 48000 | 0.4918          | {'precision': 0.6680455015511892, 'recall': 0.757327080890973, 'f1': 0.7098901098901098, 'number': 6824}  | {'precision': 0.7944098700731521, 'recall': 0.8265364080427127, 'f1': 0.8101547711836098, 'number': 8803} | {'precision': 0.9830884391461048, 'recall': 0.9893973214285714, 'f1': 0.9862327909887358, 'number': 7168} | {'precision': 0.5625, 'recall': 0.8780487804878049, 'f1': 0.6857142857142857, 'number': 82}              | {'precision': 0.9305304647725406, 'recall': 0.9845351867940921, 'f1': 0.9567713610266801, 'number': 5755} | 0.8334            | 0.8827         | 0.8574     | 0.8347           |
+| 0.3987        | 9.7   | 49000 | 0.4936          | {'precision': 0.6657629744384198, 'recall': 0.7557151230949589, 'f1': 0.7078929306794783, 'number': 6824} | {'precision': 0.7955613862468569, 'recall': 0.8266500056798819, 'f1': 0.8108077994428969, 'number': 8803} | {'precision': 0.9829474559822543, 'recall': 0.9891183035714286, 'f1': 0.9860232250886588, 'number': 7168} | {'precision': 0.5625, 'recall': 0.8780487804878049, 'f1': 0.6857142857142857, 'number': 82}              | {'precision': 0.9306605323693723, 'recall': 0.9841876629018245, 'f1': 0.9566759564225994, 'number': 5755} | 0.8331            | 0.8822         | 0.8570     | 0.8339           |
+| 0.4109        | 9.9   | 50000 | 0.4929          | {'precision': 0.6672281974860698, 'recall': 0.7545427901524033, 'f1': 0.7082043875937006, 'number': 6824} | {'precision': 0.794249480704056, 'recall': 0.8252868340338521, 'f1': 0.8094707520891365, 'number': 8803}  | {'precision': 0.983222407099279, 'recall': 0.9892578125, 'f1': 0.986230876216968, 'number': 7168}         | {'precision': 0.5669291338582677, 'recall': 0.8780487804878049, 'f1': 0.6889952153110047, 'number': 82}  | {'precision': 0.9305076392311483, 'recall': 0.9841876629018245, 'f1': 0.956595169734842, 'number': 5755}  | 0.8333            | 0.8816         | 0.8568     | 0.8339           |
+
+
+### Framework versions
+
+- Transformers 4.34.1
+- Pytorch 2.1.0+cu118
+- Datasets 2.14.5
+- Tokenizers 0.14.1

all_results.json

@@ -0,0 +1,41 @@
+{
+    "epoch": 10.0,
+    "eval_COMMENT": {
+        "f1": 0.6934239769904563,
+        "number": 7028,
+        "precision": 0.641354292623942,
+        "recall": 0.7546955036994878
+    },
+    "eval_NAME": {
+        "f1": 0.8038676952340695,
+        "number": 9278,
+        "precision": 0.7884535655058044,
+        "recall": 0.8198965294244449
+    },
+    "eval_QTY": {
+        "f1": 0.9885011895321174,
+        "number": 7571,
+        "precision": 0.9891548736939558,
+        "recall": 0.9878483687755911
+    },
+    "eval_RANGE_END": {
+        "f1": 0.782122905027933,
+        "number": 93,
+        "precision": 0.813953488372093,
+        "recall": 0.7526881720430108
+    },
+    "eval_UNIT": {
+        "f1": 0.9542504607002644,
+        "number": 6056,
+        "precision": 0.9268482490272374,
+        "recall": 0.9833223249669749
+    },
+    "eval_loss": 0.48259130120277405,
+    "eval_overall_accuracy": 0.8288907840540001,
+    "eval_overall_f1": 0.8518769349845202,
+    "eval_overall_precision": 0.8257267896217568,
+    "eval_overall_recall": 0.8797375607806568,
+    "eval_runtime": 9.3518,
+    "eval_samples_per_second": 957.465,
+    "eval_steps_per_second": 29.941
+}

config.json

@@ -0,0 +1,52 @@
+{
+  "_name_or_path": "sentence-transformers/paraphrase-MiniLM-L3-v2",
+  "architectures": [
+    "PretrainedTaggerModel"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "classifier_dropout": null,
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 384,
+  "id2label": {
+    "0": "O",
+    "1": "B-COMMENT",
+    "2": "I-COMMENT",
+    "3": "B-NAME",
+    "4": "I-NAME",
+    "5": "B-RANGE_END",
+    "6": "I-RANGE_END",
+    "7": "B-QTY",
+    "8": "I-QTY",
+    "9": "B-UNIT",
+    "10": "I-UNIT"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 1536,
+  "label2id": {
+    "B-COMMENT": 1,
+    "B-NAME": 3,
+    "B-QTY": 7,
+    "B-RANGE_END": 5,
+    "B-UNIT": 9,
+    "I-COMMENT": 2,
+    "I-NAME": 4,
+    "I-QTY": 8,
+    "I-RANGE_END": 6,
+    "I-UNIT": 10,
+    "O": 0
+  },
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 3,
+  "pad_token_id": 0,
+  "position_embedding_type": "absolute",
+  "torch_dtype": "float32",
+  "transformers_version": "4.34.1",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 30522
+}

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:712bbc1ac2a3628667510b7fd6d17583fe41c869b920714f4e0b49bed5d2a1e6
+size 69004115

special_tokens_map.json

@@ -0,0 +1,7 @@
+{
+  "cls_token": "[CLS]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
+}

test_results.json

@@ -0,0 +1,41 @@
+{
+    "epoch": 10.0,
+    "eval_COMMENT": {
+        "f1": 0.6934239769904563,
+        "number": 7028,
+        "precision": 0.641354292623942,
+        "recall": 0.7546955036994878
+    },
+    "eval_NAME": {
+        "f1": 0.8038676952340695,
+        "number": 9278,
+        "precision": 0.7884535655058044,
+        "recall": 0.8198965294244449
+    },
+    "eval_QTY": {
+        "f1": 0.9885011895321174,
+        "number": 7571,
+        "precision": 0.9891548736939558,
+        "recall": 0.9878483687755911
+    },
+    "eval_RANGE_END": {
+        "f1": 0.782122905027933,
+        "number": 93,
+        "precision": 0.813953488372093,
+        "recall": 0.7526881720430108
+    },
+    "eval_UNIT": {
+        "f1": 0.9542504607002644,
+        "number": 6056,
+        "precision": 0.9268482490272374,
+        "recall": 0.9833223249669749
+    },
+    "eval_loss": 0.48259130120277405,
+    "eval_overall_accuracy": 0.8288907840540001,
+    "eval_overall_f1": 0.8518769349845202,
+    "eval_overall_precision": 0.8257267896217568,
+    "eval_overall_recall": 0.8797375607806568,
+    "eval_runtime": 9.3518,
+    "eval_samples_per_second": 957.465,
+    "eval_steps_per_second": 29.941
+}

tokenizer_config.json

@@ -0,0 +1,64 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "[PAD]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "100": {
+      "content": "[UNK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "101": {
+      "content": "[CLS]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "102": {
+      "content": "[SEP]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "103": {
+      "content": "[MASK]",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "clean_up_tokenization_spaces": true,
+  "cls_token": "[CLS]",
+  "do_basic_tokenize": true,
+  "do_lower_case": true,
+  "mask_token": "[MASK]",
+  "max_length": 128,
+  "model_max_length": 512,
+  "never_split": null,
+  "pad_to_multiple_of": null,
+  "pad_token": "[PAD]",
+  "pad_token_type_id": 0,
+  "padding_side": "right",
+  "sep_token": "[SEP]",
+  "stride": 0,
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "BertTokenizer",
+  "truncation_side": "right",
+  "truncation_strategy": "longest_first",
+  "unk_token": "[UNK]"
+}

train_results.json

@@ -0,0 +1,41 @@
+{
+    "epoch": 10.0,
+    "eval_COMMENT": {
+        "f1": 0.7417080532947575,
+        "number": 129936,
+        "precision": 0.6935201987397969,
+        "recall": 0.7970924147272503
+    },
+    "eval_NAME": {
+        "f1": 0.8212697046102124,
+        "number": 167719,
+        "precision": 0.8069888698334503,
+        "recall": 0.8360650850529755
+    },
+    "eval_QTY": {
+        "f1": 0.9876099415118523,
+        "number": 135725,
+        "precision": 0.9873990116583078,
+        "recall": 0.9878209615030392
+    },
+    "eval_RANGE_END": {
+        "f1": 0.7593140153755175,
+        "number": 1689,
+        "precision": 0.7584170112226817,
+        "recall": 0.7602131438721137
+    },
+    "eval_UNIT": {
+        "f1": 0.9566528964458508,
+        "number": 108742,
+        "precision": 0.9318580135499228,
+        "recall": 0.9828033326589543
+    },
+    "eval_loss": 0.41290220618247986,
+    "eval_overall_accuracy": 0.8548593048477012,
+    "eval_overall_f1": 0.8686140453248525,
+    "eval_overall_precision": 0.8448665486922572,
+    "eval_overall_recall": 0.8937351395981324,
+    "eval_runtime": 128.938,
+    "eval_samples_per_second": 1253.339,
+    "eval_steps_per_second": 39.174
+}

trainer_state.json

@@ -0,0 +1,2734 @@
+{
+  "best_metric": 0.47966957092285156,
+  "best_model_checkpoint": "nyt_ingredients-tagger-paraphrase-MiniLM-L3-v2/checkpoint-24000",
+  "epoch": 10.0,
+  "eval_steps": 1000,
+  "global_step": 50510,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9505048505246485e-05,
+      "loss": 1.1204,
+      "step": 500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.9010097010492975e-05,
+      "loss": 0.7174,
+      "step": 1000
+    },
+    {
+      "epoch": 0.2,
+      "eval_COMMENT": {
+        "f1": 0.5647948868453662,
+        "number": 6824,
+        "precision": 0.5088729580444236,
+        "recall": 0.6345252051582649
+      },
+      "eval_NAME": {
+        "f1": 0.7765338110165697,
+        "number": 8803,
+        "precision": 0.7654784240150094,
+        "recall": 0.7879132114052028
+      },
+      "eval_QTY": {
+        "f1": 0.9714048901782015,
+        "number": 7168,
+        "precision": 0.9619699042407661,
+        "recall": 0.9810267857142857
+      },
+      "eval_RANGE_END": {
+        "f1": 0.0,
+        "number": 82,
+        "precision": 0.0,
+        "recall": 0.0
+      },
+      "eval_UNIT": {
+        "f1": 0.945750755794424,
+        "number": 5755,
+        "precision": 0.9151633349585568,
+        "recall": 0.9784535186794092
+      },
+      "eval_loss": 0.656587541103363,
+      "eval_overall_accuracy": 0.7841206721853672,
+      "eval_overall_f1": 0.8021117908321461,
+      "eval_overall_precision": 0.7710823961589276,
+      "eval_overall_recall": 0.8357432243643476,
+      "eval_runtime": 7.1565,
+      "eval_samples_per_second": 1188.573,
+      "eval_steps_per_second": 37.169,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.851514551573946e-05,
+      "loss": 0.6502,
+      "step": 1500
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.802019402098595e-05,
+      "loss": 0.6076,
+      "step": 2000
+    },
+    {
+      "epoch": 0.4,
+      "eval_COMMENT": {
+        "f1": 0.5854632587859425,
+        "number": 6824,
+        "precision": 0.5363414634146342,
+        "recall": 0.6444900351699883
+      },
+      "eval_NAME": {
+        "f1": 0.7806724071738058,
+        "number": 8803,
+        "precision": 0.7547189819724284,
+        "recall": 0.8084743837328183
+      },
+      "eval_QTY": {
+        "f1": 0.9778333680772705,
+        "number": 7168,
+        "precision": 0.974110480409802,
+        "recall": 0.9815848214285714
+      },
+      "eval_RANGE_END": {
+        "f1": 0.23529411764705882,
+        "number": 82,
+        "precision": 0.6,
+        "recall": 0.14634146341463414
+      },
+      "eval_UNIT": {
+        "f1": 0.9454667112746737,
+        "number": 5755,
+        "precision": 0.9114658925979681,
+        "recall": 0.9821025195482189
+      },
+      "eval_loss": 0.584534227848053,
+      "eval_overall_accuracy": 0.7988032194543752,
+      "eval_overall_f1": 0.811141258834958,
+      "eval_overall_precision": 0.7792688421188132,
+      "eval_overall_recall": 0.8457320480581167,
+      "eval_runtime": 6.3972,
+      "eval_samples_per_second": 1329.645,
+      "eval_steps_per_second": 41.581,
+      "step": 2000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.752524252623243e-05,
+      "loss": 0.5843,
+      "step": 2500
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.703029103147892e-05,
+      "loss": 0.5777,
+      "step": 3000
+    },
+    {
+      "epoch": 0.59,
+      "eval_COMMENT": {
+        "f1": 0.6086500655307995,
+        "number": 6824,
+        "precision": 0.55049786628734,
+        "recall": 0.6805392731535757
+      },
+      "eval_NAME": {
+        "f1": 0.7919685823330935,
+        "number": 8803,
+        "precision": 0.7717766278568349,
+        "recall": 0.8132454844939225
+      },
+      "eval_QTY": {
+        "f1": 0.9790267875683533,
+        "number": 7168,
+        "precision": 0.9715620277510647,
+        "recall": 0.9866071428571429
+      },
+      "eval_RANGE_END": {
+        "f1": 0.5133689839572192,
+        "number": 82,
+        "precision": 0.45714285714285713,
+        "recall": 0.5853658536585366
+      },
+      "eval_UNIT": {
+        "f1": 0.9490924440692275,
+        "number": 5755,
+        "precision": 0.9229885057471264,
+        "recall": 0.9767158992180712
+      },
+      "eval_loss": 0.5503401756286621,
+      "eval_overall_accuracy": 0.805647427686543,
+      "eval_overall_f1": 0.820622555083754,
+      "eval_overall_precision": 0.7870198165843648,
+      "eval_overall_recall": 0.8572226879016485,
+      "eval_runtime": 8.9283,
+      "eval_samples_per_second": 952.704,
+      "eval_steps_per_second": 29.793,
+      "step": 3000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.65353395367254e-05,
+      "loss": 0.5581,
+      "step": 3500
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.6040388041971886e-05,
+      "loss": 0.553,
+      "step": 4000
+    },
+    {
+      "epoch": 0.79,
+      "eval_COMMENT": {
+        "f1": 0.6282245827010623,
+        "number": 6824,
+        "precision": 0.5713428537141486,
+        "recall": 0.6976846424384525
+      },
+      "eval_NAME": {
+        "f1": 0.7973033207042568,
+        "number": 8803,
+        "precision": 0.7823947512301804,
+        "recall": 0.812791093945246
+      },
+      "eval_QTY": {
+        "f1": 0.9824755835699939,
+        "number": 7168,
+        "precision": 0.9756500206355757,
+        "recall": 0.9893973214285714
+      },
+      "eval_RANGE_END": {
+        "f1": 0.5595238095238095,
+        "number": 82,
+        "precision": 0.5465116279069767,
+        "recall": 0.573170731707317
+      },
+      "eval_UNIT": {
+        "f1": 0.9491468718634996,
+        "number": 5755,
+        "precision": 0.9150137074665377,
+        "recall": 0.9859252823631625
+      },
+      "eval_loss": 0.5342816710472107,
+      "eval_overall_accuracy": 0.8121666316171832,
+      "eval_overall_f1": 0.8289142895451346,
+      "eval_overall_precision": 0.7968357285557776,
+      "eval_overall_recall": 0.8636839899413243,
+      "eval_runtime": 7.0508,
+      "eval_samples_per_second": 1206.385,
+      "eval_steps_per_second": 37.726,
+      "step": 4000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 4.5545436547218376e-05,
+      "loss": 0.5436,
+      "step": 4500
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.505048505246486e-05,
+      "loss": 0.5407,
+      "step": 5000
+    },
+    {
+      "epoch": 0.99,
+      "eval_COMMENT": {
+        "f1": 0.6475744396119103,
+        "number": 6824,
+        "precision": 0.5958625785001848,
+        "recall": 0.7091148886283705
+      },
+      "eval_NAME": {
+        "f1": 0.8001564333202973,
+        "number": 8803,
+        "precision": 0.787269129287599,
+        "recall": 0.8134726797682609
+      },
+      "eval_QTY": {
+        "f1": 0.9841955023323818,
+        "number": 7168,
+        "precision": 0.9823488533703961,
+        "recall": 0.9860491071428571
+      },
+      "eval_RANGE_END": {
+        "f1": 0.626865671641791,
+        "number": 82,
+        "precision": 0.5294117647058824,
+        "recall": 0.7682926829268293
+      },
+      "eval_UNIT": {
+        "f1": 0.9520282780676653,
+        "number": 5755,
+        "precision": 0.9231271421576628,
+        "recall": 0.9827975673327541
+      },
+      "eval_loss": 0.518293023109436,
+      "eval_overall_accuracy": 0.8185520102471945,
+      "eval_overall_f1": 0.8361275088547816,
+      "eval_overall_precision": 0.8085002283253963,
+      "eval_overall_recall": 0.8657096954456552,
+      "eval_runtime": 7.4009,
+      "eval_samples_per_second": 1149.315,
+      "eval_steps_per_second": 35.941,
+      "step": 5000
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 4.455553355771135e-05,
+      "loss": 0.5231,
+      "step": 5500
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 4.406058206295783e-05,
+      "loss": 0.5109,
+      "step": 6000
+    },
+    {
+      "epoch": 1.19,
+      "eval_COMMENT": {
+        "f1": 0.6651526284149355,
+        "number": 6824,
+        "precision": 0.611145206824598,
+        "recall": 0.729630715123095
+      },
+      "eval_NAME": {
+        "f1": 0.803545545768759,
+        "number": 8803,
+        "precision": 0.788943623426382,
+        "recall": 0.8186981710780415
+      },
+      "eval_QTY": {
+        "f1": 0.9848337275636566,
+        "number": 7168,
+        "precision": 0.9822370247016375,
+        "recall": 0.9874441964285714
+      },
+      "eval_RANGE_END": {
+        "f1": 0.653061224489796,
+        "number": 82,
+        "precision": 0.5614035087719298,
+        "recall": 0.7804878048780488
+      },
+      "eval_UNIT": {
+        "f1": 0.9527340129749768,
+        "number": 5755,
+        "precision": 0.924762839385018,
+        "recall": 0.9824500434404866
+      },
+      "eval_loss": 0.5209813714027405,
+      "eval_overall_accuracy": 0.8200623243542929,
+      "eval_overall_f1": 0.841881782031408,
+      "eval_overall_precision": 0.8133220471415549,
+      "eval_overall_recall": 0.8725202570550433,
+      "eval_runtime": 8.4151,
+      "eval_samples_per_second": 1010.797,
+      "eval_steps_per_second": 31.61,
+      "step": 6000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 4.356563056820432e-05,
+      "loss": 0.5232,
+      "step": 6500
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 4.3070679073450804e-05,
+      "loss": 0.5125,
+      "step": 7000
+    },
+    {
+      "epoch": 1.39,
+      "eval_COMMENT": {
+        "f1": 0.6596956900596554,
+        "number": 6824,
+        "precision": 0.607906114885732,
+        "recall": 0.7211313012895663
+      },
+      "eval_NAME": {
+        "f1": 0.8034675615212529,
+        "number": 8803,
+        "precision": 0.7913407513495648,
+        "recall": 0.8159718277859821
+      },
+      "eval_QTY": {
+        "f1": 0.9845339278249966,
+        "number": 7168,
+        "precision": 0.983300862788756,
+        "recall": 0.9857700892857143
+      },
+      "eval_RANGE_END": {
+        "f1": 0.641711229946524,
+        "number": 82,
+        "precision": 0.5714285714285714,
+        "recall": 0.7317073170731707
+      },
+      "eval_UNIT": {
+        "f1": 0.952638700947226,
+        "number": 5755,
+        "precision": 0.9279947273026857,
+        "recall": 0.978627280625543
+      },
+      "eval_loss": 0.5072416663169861,
+      "eval_overall_accuracy": 0.8215152847611218,
+      "eval_overall_f1": 0.8404435129470623,
+      "eval_overall_precision": 0.8142932005764444,
+      "eval_overall_recall": 0.8683291422184968,
+      "eval_runtime": 6.3336,
+      "eval_samples_per_second": 1342.988,
+      "eval_steps_per_second": 41.998,
+      "step": 7000
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 4.257572757869729e-05,
+      "loss": 0.5084,
+      "step": 7500
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 4.208077608394378e-05,
+      "loss": 0.4967,
+      "step": 8000
+    },
+    {
+      "epoch": 1.58,
+      "eval_COMMENT": {
+        "f1": 0.6653271716562856,
+        "number": 6824,
+        "precision": 0.6126803996546195,
+        "recall": 0.7278722157092614
+      },
+      "eval_NAME": {
+        "f1": 0.8041098197742967,
+        "number": 8803,
+        "precision": 0.7949600355239786,
+        "recall": 0.8134726797682609
+      },
+      "eval_QTY": {
+        "f1": 0.9852011394427846,
+        "number": 7168,
+        "precision": 0.9813148788927336,
+        "recall": 0.9891183035714286
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6431718061674009,
+        "number": 82,
+        "precision": 0.503448275862069,
+        "recall": 0.8902439024390244
+      },
+      "eval_UNIT": {
+        "f1": 0.9540510918135063,
+        "number": 5755,
+        "precision": 0.9266295447101212,
+        "recall": 0.9831450912250217
+      },
+      "eval_loss": 0.5088181495666504,
+      "eval_overall_accuracy": 0.8218402890626494,
+      "eval_overall_f1": 0.8425442817824156,
+      "eval_overall_precision": 0.815566669935602,
+      "eval_overall_recall": 0.871367700474993,
+      "eval_runtime": 8.617,
+      "eval_samples_per_second": 987.123,
+      "eval_steps_per_second": 30.869,
+      "step": 8000
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 4.158582458919026e-05,
+      "loss": 0.5024,
+      "step": 8500
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 4.109087309443675e-05,
+      "loss": 0.5121,
+      "step": 9000
+    },
+    {
+      "epoch": 1.78,
+      "eval_COMMENT": {
+        "f1": 0.6751009421265142,
+        "number": 6824,
+        "precision": 0.6241911398705824,
+        "recall": 0.735052754982415
+      },
+      "eval_NAME": {
+        "f1": 0.803609223571349,
+        "number": 8803,
+        "precision": 0.7883291443558081,
+        "recall": 0.8194933545382256
+      },
+      "eval_QTY": {
+        "f1": 0.9847892827239743,
+        "number": 7168,
+        "precision": 0.9850642099385818,
+        "recall": 0.9845145089285714
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6346153846153846,
+        "number": 82,
+        "precision": 0.5238095238095238,
+        "recall": 0.8048780487804879
+      },
+      "eval_UNIT": {
+        "f1": 0.9542967763656258,
+        "number": 5755,
+        "precision": 0.9253999347045381,
+        "recall": 0.9850564726324935
+      },
+      "eval_loss": 0.502194344997406,
+      "eval_overall_accuracy": 0.824765327776397,
+      "eval_overall_f1": 0.8448383557018655,
+      "eval_overall_precision": 0.8176322582753325,
+      "eval_overall_recall": 0.8739172953338922,
+      "eval_runtime": 8.4909,
+      "eval_samples_per_second": 1001.783,
+      "eval_steps_per_second": 31.328,
+      "step": 9000
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 4.059592159968323e-05,
+      "loss": 0.4946,
+      "step": 9500
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 4.010097010492972e-05,
+      "loss": 0.5008,
+      "step": 10000
+    },
+    {
+      "epoch": 1.98,
+      "eval_COMMENT": {
+        "f1": 0.6771771771771772,
+        "number": 6824,
+        "precision": 0.6337506387327542,
+        "recall": 0.7269929660023446
+      },
+      "eval_NAME": {
+        "f1": 0.8049774008146867,
+        "number": 8803,
+        "precision": 0.7910726036411494,
+        "recall": 0.8193797569010565
+      },
+      "eval_QTY": {
+        "f1": 0.9824440619621343,
+        "number": 7168,
+        "precision": 0.9698246567894522,
+        "recall": 0.9953962053571429
+      },
+      "eval_RANGE_END": {
+        "f1": 0.625,
+        "number": 82,
+        "precision": 0.47468354430379744,
+        "recall": 0.9146341463414634
+      },
+      "eval_UNIT": {
+        "f1": 0.9544190749010025,
+        "number": 5755,
+        "precision": 0.9263984298331698,
+        "recall": 0.9841876629018245
+      },
+      "eval_loss": 0.4997539520263672,
+      "eval_overall_accuracy": 0.8250138604775652,
+      "eval_overall_f1": 0.8461161686962692,
+      "eval_overall_precision": 0.8192313982011448,
+      "eval_overall_recall": 0.874825370215144,
+      "eval_runtime": 6.4794,
+      "eval_samples_per_second": 1312.774,
+      "eval_steps_per_second": 41.053,
+      "step": 10000
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 3.9606018610176205e-05,
+      "loss": 0.4788,
+      "step": 10500
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.911106711542269e-05,
+      "loss": 0.4863,
+      "step": 11000
+    },
+    {
+      "epoch": 2.18,
+      "eval_COMMENT": {
+        "f1": 0.6816733337853415,
+        "number": 6824,
+        "precision": 0.6343217665615142,
+        "recall": 0.736664712778429
+      },
+      "eval_NAME": {
+        "f1": 0.8056887897378694,
+        "number": 8803,
+        "precision": 0.7913881888901063,
+        "recall": 0.820515733272748
+      },
+      "eval_QTY": {
+        "f1": 0.9852848873701096,
+        "number": 7168,
+        "precision": 0.9850787895690978,
+        "recall": 0.9854910714285714
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6666666666666666,
+        "number": 82,
+        "precision": 0.6122448979591837,
+        "recall": 0.7317073170731707
+      },
+      "eval_UNIT": {
+        "f1": 0.9536959084020878,
+        "number": 5755,
+        "precision": 0.9250367466927977,
+        "recall": 0.9841876629018245
+      },
+      "eval_loss": 0.49540793895721436,
+      "eval_overall_accuracy": 0.8272888905882578,
+      "eval_overall_f1": 0.8476538695917124,
+      "eval_overall_precision": 0.8224280646432794,
+      "eval_overall_recall": 0.8744761106454317,
+      "eval_runtime": 9.0889,
+      "eval_samples_per_second": 935.867,
+      "eval_steps_per_second": 29.266,
+      "step": 11000
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 3.861611562066917e-05,
+      "loss": 0.4922,
+      "step": 11500
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 3.812116412591566e-05,
+      "loss": 0.4893,
+      "step": 12000
+    },
+    {
+      "epoch": 2.38,
+      "eval_COMMENT": {
+        "f1": 0.6847270771712826,
+        "number": 6824,
+        "precision": 0.6325136612021858,
+        "recall": 0.7463364595545134
+      },
+      "eval_NAME": {
+        "f1": 0.8033142412278262,
+        "number": 8803,
+        "precision": 0.7868191721132898,
+        "recall": 0.820515733272748
+      },
+      "eval_QTY": {
+        "f1": 0.9853482397055761,
+        "number": 7168,
+        "precision": 0.9809207797594359,
+        "recall": 0.9898158482142857
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6491228070175438,
+        "number": 82,
+        "precision": 0.5068493150684932,
+        "recall": 0.9024390243902439
+      },
+      "eval_UNIT": {
+        "f1": 0.9545187053383775,
+        "number": 5755,
+        "precision": 0.9245928338762215,
+        "recall": 0.9864465682015638
+      },
+      "eval_loss": 0.49466800689697266,
+      "eval_overall_accuracy": 0.8257212227808898,
+      "eval_overall_f1": 0.8474479228061903,
+      "eval_overall_precision": 0.818249813014211,
+      "eval_overall_recall": 0.878806929309863,
+      "eval_runtime": 7.6267,
+      "eval_samples_per_second": 1115.288,
+      "eval_steps_per_second": 34.877,
+      "step": 12000
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 3.762621263116215e-05,
+      "loss": 0.4767,
+      "step": 12500
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 3.7131261136408633e-05,
+      "loss": 0.489,
+      "step": 13000
+    },
+    {
+      "epoch": 2.57,
+      "eval_COMMENT": {
+        "f1": 0.6822220714915552,
+        "number": 6824,
+        "precision": 0.635054931178179,
+        "recall": 0.736957796014068
+      },
+      "eval_NAME": {
+        "f1": 0.8048549635320973,
+        "number": 8803,
+        "precision": 0.7892552959161389,
+        "recall": 0.8210837214585937
+      },
+      "eval_QTY": {
+        "f1": 0.9860937282714505,
+        "number": 7168,
+        "precision": 0.9829498197948433,
+        "recall": 0.9892578125
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6574074074074074,
+        "number": 82,
+        "precision": 0.5298507462686567,
+        "recall": 0.8658536585365854
+      },
+      "eval_UNIT": {
+        "f1": 0.9543087518898035,
+        "number": 5755,
+        "precision": 0.9235896602178507,
+        "recall": 0.9871416159860991
+      },
+      "eval_loss": 0.48692840337753296,
+      "eval_overall_accuracy": 0.8271550652876288,
+      "eval_overall_f1": 0.8478583975138494,
+      "eval_overall_precision": 0.8209052851909995,
+      "eval_overall_recall": 0.8766415199776474,
+      "eval_runtime": 6.7601,
+      "eval_samples_per_second": 1258.259,
+      "eval_steps_per_second": 39.348,
+      "step": 13000
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 3.663630964165512e-05,
+      "loss": 0.4794,
+      "step": 13500
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.6141358146901606e-05,
+      "loss": 0.4691,
+      "step": 14000
+    },
+    {
+      "epoch": 2.77,
+      "eval_COMMENT": {
+        "f1": 0.6935899173329719,
+        "number": 6824,
+        "precision": 0.6450718427022939,
+        "recall": 0.75
+      },
+      "eval_NAME": {
+        "f1": 0.8057417336907953,
+        "number": 8803,
+        "precision": 0.7925502692011867,
+        "recall": 0.8193797569010565
+      },
+      "eval_QTY": {
+        "f1": 0.9854750156369451,
+        "number": 7168,
+        "precision": 0.9818584683561834,
+        "recall": 0.9891183035714286
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6820276497695853,
+        "number": 82,
+        "precision": 0.5481481481481482,
+        "recall": 0.9024390243902439
+      },
+      "eval_UNIT": {
+        "f1": 0.9554861052453754,
+        "number": 5755,
+        "precision": 0.9296515450361604,
+        "recall": 0.9827975673327541
+      },
+      "eval_loss": 0.4904623329639435,
+      "eval_overall_accuracy": 0.8282639034928403,
+      "eval_overall_f1": 0.8510328725870033,
+      "eval_overall_precision": 0.8252994257588186,
+      "eval_overall_recall": 0.8784227437831796,
+      "eval_runtime": 9.0651,
+      "eval_samples_per_second": 938.328,
+      "eval_steps_per_second": 29.343,
+      "step": 14000
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 3.564640665214809e-05,
+      "loss": 0.484,
+      "step": 14500
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 3.515145515739457e-05,
+      "loss": 0.4781,
+      "step": 15000
+    },
+    {
+      "epoch": 2.97,
+      "eval_COMMENT": {
+        "f1": 0.6936924742826204,
+        "number": 6824,
+        "precision": 0.6444919517102615,
+        "recall": 0.7510257913247362
+      },
+      "eval_NAME": {
+        "f1": 0.8070410729253981,
+        "number": 8803,
+        "precision": 0.7942146942366916,
+        "recall": 0.8202885379984096
+      },
+      "eval_QTY": {
+        "f1": 0.9859409799554566,
+        "number": 7168,
+        "precision": 0.98375,
+        "recall": 0.9881417410714286
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6602870813397129,
+        "number": 82,
+        "precision": 0.5433070866141733,
+        "recall": 0.8414634146341463
+      },
+      "eval_UNIT": {
+        "f1": 0.9547484621218505,
+        "number": 5755,
+        "precision": 0.9268651832460733,
+        "recall": 0.9843614248479583
+      },
+      "eval_loss": 0.48784932494163513,
+      "eval_overall_accuracy": 0.8293345058978722,
+      "eval_overall_f1": 0.8513236911105472,
+      "eval_overall_precision": 0.825476495095627,
+      "eval_overall_recall": 0.8788418552668343,
+      "eval_runtime": 7.0221,
+      "eval_samples_per_second": 1211.315,
+      "eval_steps_per_second": 37.88,
+      "step": 15000
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 3.465650366264107e-05,
+      "loss": 0.462,
+      "step": 15500
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 3.416155216788755e-05,
+      "loss": 0.4599,
+      "step": 16000
+    },
+    {
+      "epoch": 3.17,
+      "eval_COMMENT": {
+        "f1": 0.6904161841836387,
+        "number": 6824,
+        "precision": 0.6450299096347206,
+        "recall": 0.742672919109027
+      },
+      "eval_NAME": {
+        "f1": 0.8082375265096551,
+        "number": 8803,
+        "precision": 0.7944048272078991,
+        "recall": 0.8225604907417926
+      },
+      "eval_QTY": {
+        "f1": 0.9857391304347826,
+        "number": 7168,
+        "precision": 0.983072013320383,
+        "recall": 0.9884207589285714
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6636363636363637,
+        "number": 82,
+        "precision": 0.5289855072463768,
+        "recall": 0.8902439024390244
+      },
+      "eval_UNIT": {
+        "f1": 0.9555706291334577,
+        "number": 5755,
+        "precision": 0.9331015068719987,
+        "recall": 0.9791485664639444
+      },
+      "eval_loss": 0.48515409231185913,
+      "eval_overall_accuracy": 0.8305962873038025,
+      "eval_overall_f1": 0.8510883569539567,
+      "eval_overall_precision": 0.8269205428910265,
+      "eval_overall_recall": 0.8767113718915899,
+      "eval_runtime": 8.5334,
+      "eval_samples_per_second": 996.787,
+      "eval_steps_per_second": 31.172,
+      "step": 16000
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 3.3666600673134034e-05,
+      "loss": 0.4548,
+      "step": 16500
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 3.317164917838052e-05,
+      "loss": 0.4636,
+      "step": 17000
+    },
+    {
+      "epoch": 3.37,
+      "eval_COMMENT": {
+        "f1": 0.6982570806100218,
+        "number": 6824,
+        "precision": 0.6520854526958291,
+        "recall": 0.7514654161781946
+      },
+      "eval_NAME": {
+        "f1": 0.8060380709251347,
+        "number": 8803,
+        "precision": 0.7879774305555556,
+        "recall": 0.8249460411223447
+      },
+      "eval_QTY": {
+        "f1": 0.9860387580745988,
+        "number": 7168,
+        "precision": 0.9818785447503112,
+        "recall": 0.990234375
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6198347107438017,
+        "number": 82,
+        "precision": 0.46875,
+        "recall": 0.9146341463414634
+      },
+      "eval_UNIT": {
+        "f1": 0.9546639751030365,
+        "number": 5755,
+        "precision": 0.9251711770459733,
+        "recall": 0.9860990443092963
+      },
+      "eval_loss": 0.4889169931411743,
+      "eval_overall_accuracy": 0.8302904009023648,
+      "eval_overall_f1": 0.8521313412678315,
+      "eval_overall_precision": 0.8246903898310624,
+      "eval_overall_recall": 0.8814613020396759,
+      "eval_runtime": 8.1836,
+      "eval_samples_per_second": 1039.392,
+      "eval_steps_per_second": 32.504,
+      "step": 17000
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 3.267669768362701e-05,
+      "loss": 0.4638,
+      "step": 17500
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 3.218174618887349e-05,
+      "loss": 0.4688,
+      "step": 18000
+    },
+    {
+      "epoch": 3.56,
+      "eval_COMMENT": {
+        "f1": 0.7010448264239973,
+        "number": 6824,
+        "precision": 0.6491074772188241,
+        "recall": 0.7620164126611958
+      },
+      "eval_NAME": {
+        "f1": 0.8081609837898266,
+        "number": 8803,
+        "precision": 0.7955320787938813,
+        "recall": 0.8211973190957628
+      },
+      "eval_QTY": {
+        "f1": 0.9860232250886588,
+        "number": 7168,
+        "precision": 0.9829474559822543,
+        "recall": 0.9891183035714286
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6428571428571429,
+        "number": 82,
+        "precision": 0.5526315789473685,
+        "recall": 0.7682926829268293
+      },
+      "eval_UNIT": {
+        "f1": 0.9546716003700276,
+        "number": 5755,
+        "precision": 0.9250325945241199,
+        "recall": 0.98627280625543
+      },
+      "eval_loss": 0.49013254046440125,
+      "eval_overall_accuracy": 0.8319345403100923,
+      "eval_overall_f1": 0.8534117209805214,
+      "eval_overall_precision": 0.8264781911586663,
+      "eval_overall_recall": 0.8821598211791003,
+      "eval_runtime": 7.2351,
+      "eval_samples_per_second": 1175.656,
+      "eval_steps_per_second": 36.765,
+      "step": 18000
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 3.168679469411997e-05,
+      "loss": 0.4648,
+      "step": 18500
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 3.119184319936647e-05,
+      "loss": 0.4584,
+      "step": 19000
+    },
+    {
+      "epoch": 3.76,
+      "eval_COMMENT": {
+        "f1": 0.6999050332383666,
+        "number": 6824,
+        "precision": 0.65155342258146,
+        "recall": 0.7560082063305978
+      },
+      "eval_NAME": {
+        "f1": 0.8059467918622849,
+        "number": 8803,
+        "precision": 0.7932665859830564,
+        "recall": 0.819038963989549
+      },
+      "eval_QTY": {
+        "f1": 0.9858998402444954,
+        "number": 7168,
+        "precision": 0.981740213030848,
+        "recall": 0.9900948660714286
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6636363636363637,
+        "number": 82,
+        "precision": 0.5289855072463768,
+        "recall": 0.8902439024390244
+      },
+      "eval_UNIT": {
+        "f1": 0.956632006086736,
+        "number": 5755,
+        "precision": 0.9315113598946329,
+        "recall": 0.9831450912250217
+      },
+      "eval_loss": 0.48241040110588074,
+      "eval_overall_accuracy": 0.8306727589041619,
+      "eval_overall_f1": 0.8529790115098173,
+      "eval_overall_precision": 0.8275420388859696,
+      "eval_overall_recall": 0.8800293378038558,
+      "eval_runtime": 9.4362,
+      "eval_samples_per_second": 901.421,
+      "eval_steps_per_second": 28.189,
+      "step": 19000
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 3.069689170461295e-05,
+      "loss": 0.4584,
+      "step": 19500
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 3.0201940209859435e-05,
+      "loss": 0.4701,
+      "step": 20000
+    },
+    {
+      "epoch": 3.96,
+      "eval_COMMENT": {
+        "f1": 0.7003649886371462,
+        "number": 6824,
+        "precision": 0.6606470053267507,
+        "recall": 0.7451641266119577
+      },
+      "eval_NAME": {
+        "f1": 0.8087504883084994,
+        "number": 8803,
+        "precision": 0.7948661693725319,
+        "recall": 0.8231284789276383
+      },
+      "eval_QTY": {
+        "f1": 0.98635477582846,
+        "number": 7168,
+        "precision": 0.9844357976653697,
+        "recall": 0.98828125
+      },
+      "eval_RANGE_END": {
+        "f1": 0.660377358490566,
+        "number": 82,
+        "precision": 0.5384615384615384,
+        "recall": 0.8536585365853658
+      },
+      "eval_UNIT": {
+        "f1": 0.956190798376184,
+        "number": 5755,
+        "precision": 0.9314549349151425,
+        "recall": 0.9822762814943528
+      },
+      "eval_loss": 0.48059511184692383,
+      "eval_overall_accuracy": 0.8328713174144952,
+      "eval_overall_f1": 0.8544527532290959,
+      "eval_overall_precision": 0.8321636652542372,
+      "eval_overall_recall": 0.8779687063425538,
+      "eval_runtime": 6.465,
+      "eval_samples_per_second": 1315.706,
+      "eval_steps_per_second": 41.145,
+      "step": 20000
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 2.970698871510592e-05,
+      "loss": 0.4594,
+      "step": 20500
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 2.9212037220352405e-05,
+      "loss": 0.4441,
+      "step": 21000
+    },
+    {
+      "epoch": 4.16,
+      "eval_COMMENT": {
+        "f1": 0.6940426407074792,
+        "number": 6824,
+        "precision": 0.6520674996779595,
+        "recall": 0.7417936694021102
+      },
+      "eval_NAME": {
+        "f1": 0.8061428889383486,
+        "number": 8803,
+        "precision": 0.7900534409423056,
+        "recall": 0.8229012836533001
+      },
+      "eval_QTY": {
+        "f1": 0.9851016429963798,
+        "number": 7168,
+        "precision": 0.9831851028349082,
+        "recall": 0.9870256696428571
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6521739130434783,
+        "number": 82,
+        "precision": 0.5882352941176471,
+        "recall": 0.7317073170731707
+      },
+      "eval_UNIT": {
+        "f1": 0.9563670253325427,
+        "number": 5755,
+        "precision": 0.9332010582010583,
+        "recall": 0.9807124239791486
+      },
+      "eval_loss": 0.4863247871398926,
+      "eval_overall_accuracy": 0.8309977632056895,
+      "eval_overall_f1": 0.851638092004753,
+      "eval_overall_precision": 0.8284893321883876,
+      "eval_overall_recall": 0.8761176306230791,
+      "eval_runtime": 8.7832,
+      "eval_samples_per_second": 968.441,
+      "eval_steps_per_second": 30.285,
+      "step": 21000
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 2.871708572559889e-05,
+      "loss": 0.4493,
+      "step": 21500
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 2.8222134230845377e-05,
+      "loss": 0.4375,
+      "step": 22000
+    },
+    {
+      "epoch": 4.36,
+      "eval_COMMENT": {
+        "f1": 0.699911438108863,
+        "number": 6824,
+        "precision": 0.6539783577339274,
+        "recall": 0.7527842907385698
+      },
+      "eval_NAME": {
+        "f1": 0.8084585420144685,
+        "number": 8803,
+        "precision": 0.7924075488164066,
+        "recall": 0.8251732363966829
+      },
+      "eval_QTY": {
+        "f1": 0.985897881208753,
+        "number": 7168,
+        "precision": 0.9818735298187353,
+        "recall": 0.9899553571428571
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6519823788546256,
+        "number": 82,
+        "precision": 0.5103448275862069,
+        "recall": 0.9024390243902439
+      },
+      "eval_UNIT": {
+        "f1": 0.9565510841137266,
+        "number": 5755,
+        "precision": 0.9296490652673007,
+        "recall": 0.9850564726324935
+      },
+      "eval_loss": 0.4853152334690094,
+      "eval_overall_accuracy": 0.8326036668132373,
+      "eval_overall_f1": 0.8537987957513024,
+      "eval_overall_precision": 0.8277581004853732,
+      "eval_overall_recall": 0.8815311539536184,
+      "eval_runtime": 7.6535,
+      "eval_samples_per_second": 1111.391,
+      "eval_steps_per_second": 34.755,
+      "step": 22000
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 2.7727182736091867e-05,
+      "loss": 0.4541,
+      "step": 22500
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 2.7232231241338353e-05,
+      "loss": 0.4566,
+      "step": 23000
+    },
+    {
+      "epoch": 4.55,
+      "eval_COMMENT": {
+        "f1": 0.6999176728869374,
+        "number": 6824,
+        "precision": 0.6580237358101135,
+        "recall": 0.7475087924970691
+      },
+      "eval_NAME": {
+        "f1": 0.8100474462740721,
+        "number": 8803,
+        "precision": 0.7963125548726954,
+        "recall": 0.8242644552993298
+      },
+      "eval_QTY": {
+        "f1": 0.9862956521739129,
+        "number": 7168,
+        "precision": 0.9836270292770917,
+        "recall": 0.9889787946428571
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6979166666666666,
+        "number": 82,
+        "precision": 0.6090909090909091,
+        "recall": 0.8170731707317073
+      },
+      "eval_UNIT": {
+        "f1": 0.9566618607412434,
+        "number": 5755,
+        "precision": 0.9343936381709742,
+        "recall": 0.9800173761946134
+      },
+      "eval_loss": 0.48071911931037903,
+      "eval_overall_accuracy": 0.8325271952128778,
+      "eval_overall_f1": 0.8548318578055702,
+      "eval_overall_precision": 0.8324122182877188,
+      "eval_overall_recall": 0.8784925956971221,
+      "eval_runtime": 7.014,
+      "eval_samples_per_second": 1212.724,
+      "eval_steps_per_second": 37.924,
+      "step": 23000
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 2.6737279746584836e-05,
+      "loss": 0.4448,
+      "step": 23500
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 2.6242328251831323e-05,
+      "loss": 0.4546,
+      "step": 24000
+    },
+    {
+      "epoch": 4.75,
+      "eval_COMMENT": {
+        "f1": 0.7049292873815736,
+        "number": 6824,
+        "precision": 0.6631361405321622,
+        "recall": 0.7523446658851114
+      },
+      "eval_NAME": {
+        "f1": 0.8063276332646355,
+        "number": 8803,
+        "precision": 0.7910382513661203,
+        "recall": 0.8222196978302851
+      },
+      "eval_QTY": {
+        "f1": 0.9859881491808993,
+        "number": 7168,
+        "precision": 0.985369931726348,
+        "recall": 0.9866071428571429
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6783625730994152,
+        "number": 82,
+        "precision": 0.651685393258427,
+        "recall": 0.7073170731707317
+      },
+      "eval_UNIT": {
+        "f1": 0.9558001693480102,
+        "number": 5755,
+        "precision": 0.9321222130470685,
+        "recall": 0.9807124239791486
+      },
+      "eval_loss": 0.47966957092285156,
+      "eval_overall_accuracy": 0.832374252012159,
+      "eval_overall_f1": 0.8546520520010196,
+      "eval_overall_precision": 0.8322907357759904,
+      "eval_overall_recall": 0.8782481139983236,
+      "eval_runtime": 9.0205,
+      "eval_samples_per_second": 942.965,
+      "eval_steps_per_second": 29.488,
+      "step": 24000
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 2.5747376757077806e-05,
+      "loss": 0.4498,
+      "step": 24500
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 2.5252425262324292e-05,
+      "loss": 0.4462,
+      "step": 25000
+    },
+    {
+      "epoch": 4.95,
+      "eval_COMMENT": {
+        "f1": 0.6998077451249656,
+        "number": 6824,
+        "precision": 0.6583979328165375,
+        "recall": 0.7467760844079718
+      },
+      "eval_NAME": {
+        "f1": 0.8070350643586329,
+        "number": 8803,
+        "precision": 0.7887430864331417,
+        "recall": 0.8261956151312053
+      },
+      "eval_QTY": {
+        "f1": 0.9860329372524493,
+        "number": 7168,
+        "precision": 0.9822788315104527,
+        "recall": 0.9898158482142857
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6571428571428573,
+        "number": 82,
+        "precision": 0.5390625,
+        "recall": 0.8414634146341463
+      },
+      "eval_UNIT": {
+        "f1": 0.9555161644298135,
+        "number": 5755,
+        "precision": 0.9290873276428102,
+        "recall": 0.9834926151172894
+      },
+      "eval_loss": 0.48545241355895996,
+      "eval_overall_accuracy": 0.8324889594126981,
+      "eval_overall_f1": 0.8534792330103665,
+      "eval_overall_precision": 0.8286080778844889,
+      "eval_overall_recall": 0.879889633975971,
+      "eval_runtime": 6.8304,
+      "eval_samples_per_second": 1245.307,
+      "eval_steps_per_second": 38.943,
+      "step": 25000
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 2.4757473767570778e-05,
+      "loss": 0.4373,
+      "step": 25500
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 2.4262522272817265e-05,
+      "loss": 0.433,
+      "step": 26000
+    },
+    {
+      "epoch": 5.15,
+      "eval_COMMENT": {
+        "f1": 0.7011549238023645,
+        "number": 6824,
+        "precision": 0.656934306569343,
+        "recall": 0.7517584994138335
+      },
+      "eval_NAME": {
+        "f1": 0.805529234713784,
+        "number": 8803,
+        "precision": 0.7907638432917488,
+        "recall": 0.8208565261842554
+      },
+      "eval_QTY": {
+        "f1": 0.9860607750209088,
+        "number": 7168,
+        "precision": 0.9852367688022284,
+        "recall": 0.9868861607142857
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6607929515418502,
+        "number": 82,
+        "precision": 0.5172413793103449,
+        "recall": 0.9146341463414634
+      },
+      "eval_UNIT": {
+        "f1": 0.9570749915167968,
+        "number": 5755,
+        "precision": 0.935024034477043,
+        "recall": 0.9801911381407472
+      },
+      "eval_loss": 0.4933657944202423,
+      "eval_overall_accuracy": 0.8298315713002083,
+      "eval_overall_f1": 0.8533179496750768,
+      "eval_overall_precision": 0.8297640653357532,
+      "eval_overall_recall": 0.8782481139983236,
+      "eval_runtime": 8.2218,
+      "eval_samples_per_second": 1034.568,
+      "eval_steps_per_second": 32.353,
+      "step": 26000
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 2.376757077806375e-05,
+      "loss": 0.4353,
+      "step": 26500
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 2.3272619283310237e-05,
+      "loss": 0.442,
+      "step": 27000
+    },
+    {
+      "epoch": 5.35,
+      "eval_COMMENT": {
+        "f1": 0.7040608447643624,
+        "number": 6824,
+        "precision": 0.6560364464692483,
+        "recall": 0.7596717467760844
+      },
+      "eval_NAME": {
+        "f1": 0.8098804068402817,
+        "number": 8803,
+        "precision": 0.7970520294797052,
+        "recall": 0.8231284789276383
+      },
+      "eval_QTY": {
+        "f1": 0.9860037601838312,
+        "number": 7168,
+        "precision": 0.9842902822188239,
+        "recall": 0.9877232142857143
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6728110599078342,
+        "number": 82,
+        "precision": 0.5407407407407407,
+        "recall": 0.8902439024390244
+      },
+      "eval_UNIT": {
+        "f1": 0.956294296321296,
+        "number": 5755,
+        "precision": 0.9294735115630638,
+        "recall": 0.9847089487402259
+      },
+      "eval_loss": 0.48500868678092957,
+      "eval_overall_accuracy": 0.8332727933163822,
+      "eval_overall_f1": 0.855207451312447,
+      "eval_overall_precision": 0.8301005983299362,
+      "eval_overall_recall": 0.8818804135233306,
+      "eval_runtime": 8.8139,
+      "eval_samples_per_second": 965.065,
+      "eval_steps_per_second": 30.18,
+      "step": 27000
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 2.2777667788556724e-05,
+      "loss": 0.4321,
+      "step": 27500
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 2.2282716293803206e-05,
+      "loss": 0.4348,
+      "step": 28000
+    },
+    {
+      "epoch": 5.54,
+      "eval_COMMENT": {
+        "f1": 0.7028109676082602,
+        "number": 6824,
+        "precision": 0.6646636185499674,
+        "recall": 0.7456037514654161
+      },
+      "eval_NAME": {
+        "f1": 0.8111872527717422,
+        "number": 8803,
+        "precision": 0.7959763831183031,
+        "recall": 0.8269907985913894
+      },
+      "eval_QTY": {
+        "f1": 0.9861033907726514,
+        "number": 7168,
+        "precision": 0.982281284606866,
+        "recall": 0.9899553571428571
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6995073891625616,
+        "number": 82,
+        "precision": 0.5867768595041323,
+        "recall": 0.8658536585365854
+      },
+      "eval_UNIT": {
+        "f1": 0.9556492411467116,
+        "number": 5755,
+        "precision": 0.9282555282555283,
+        "recall": 0.9847089487402259
+      },
+      "eval_loss": 0.4908902645111084,
+      "eval_overall_accuracy": 0.8319536582101822,
+      "eval_overall_f1": 0.8560025813902146,
+      "eval_overall_precision": 0.8330964265644111,
+      "eval_overall_recall": 0.8802039675887119,
+      "eval_runtime": 6.4986,
+      "eval_samples_per_second": 1308.901,
+      "eval_steps_per_second": 40.932,
+      "step": 28000
+    },
+    {
+      "epoch": 5.64,
+      "learning_rate": 2.1787764799049696e-05,
+      "loss": 0.4335,
+      "step": 28500
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 2.129281330429618e-05,
+      "loss": 0.4324,
+      "step": 29000
+    },
+    {
+      "epoch": 5.74,
+      "eval_COMMENT": {
+        "f1": 0.7066996835878389,
+        "number": 6824,
+        "precision": 0.6659320715582059,
+        "recall": 0.7527842907385698
+      },
+      "eval_NAME": {
+        "f1": 0.8106571540047935,
+        "number": 8803,
+        "precision": 0.7957977675640184,
+        "recall": 0.8260820174940361
+      },
+      "eval_QTY": {
+        "f1": 0.9864479810966712,
+        "number": 7168,
+        "precision": 0.9828278631768453,
+        "recall": 0.9900948660714286
+      },
+      "eval_RANGE_END": {
+        "f1": 0.7035175879396985,
+        "number": 82,
+        "precision": 0.5982905982905983,
+        "recall": 0.8536585365853658
+      },
+      "eval_UNIT": {
+        "f1": 0.9554875976894326,
+        "number": 5755,
+        "precision": 0.9346850589995014,
+        "recall": 0.9772371850564726
+      },
+      "eval_loss": 0.4821859300136566,
+      "eval_overall_accuracy": 0.8341139809203357,
+      "eval_overall_f1": 0.8565747208484169,
+      "eval_overall_precision": 0.8342437183434304,
+      "eval_overall_recall": 0.8801341156747695,
+      "eval_runtime": 8.6529,
+      "eval_samples_per_second": 983.028,
+      "eval_steps_per_second": 30.741,
+      "step": 29000
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 2.0797861809542665e-05,
+      "loss": 0.4433,
+      "step": 29500
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 2.0302910314789152e-05,
+      "loss": 0.4375,
+      "step": 30000
+    },
+    {
+      "epoch": 5.94,
+      "eval_COMMENT": {
+        "f1": 0.70434363598816,
+        "number": 6824,
+        "precision": 0.6641568220173958,
+        "recall": 0.7497069167643611
+      },
+      "eval_NAME": {
+        "f1": 0.8090376569037656,
+        "number": 8803,
+        "precision": 0.7948914711686034,
+        "recall": 0.823696467113484
+      },
+      "eval_QTY": {
+        "f1": 0.9860154456272177,
+        "number": 7168,
+        "precision": 0.9834836918806384,
+        "recall": 0.9885602678571429
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6785714285714285,
+        "number": 82,
+        "precision": 0.5352112676056338,
+        "recall": 0.926829268292683
+      },
+      "eval_UNIT": {
+        "f1": 0.9560281284419215,
+        "number": 5755,
+        "precision": 0.9328703703703703,
+        "recall": 0.9803649000868809
+      },
+      "eval_loss": 0.48426133394241333,
+      "eval_overall_accuracy": 0.8327948458141358,
+      "eval_overall_f1": 0.8553999864065792,
+      "eval_overall_precision": 0.8329252150893448,
+      "eval_overall_recall": 0.8791212629226041,
+      "eval_runtime": 7.6434,
+      "eval_samples_per_second": 1112.862,
+      "eval_steps_per_second": 34.801,
+      "step": 30000
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 1.9807958820035638e-05,
+      "loss": 0.4331,
+      "step": 30500
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 1.9313007325282124e-05,
+      "loss": 0.4226,
+      "step": 31000
+    },
+    {
+      "epoch": 6.14,
+      "eval_COMMENT": {
+        "f1": 0.7090600610790635,
+        "number": 6824,
+        "precision": 0.6603463531791177,
+        "recall": 0.7655334114888629
+      },
+      "eval_NAME": {
+        "f1": 0.806347438752784,
+        "number": 8803,
+        "precision": 0.7907611663208475,
+        "recall": 0.8225604907417926
+      },
+      "eval_QTY": {
+        "f1": 0.9856705620478576,
+        "number": 7168,
+        "precision": 0.9829356270810211,
+        "recall": 0.9884207589285714
+      },
+      "eval_RANGE_END": {
+        "f1": 0.669603524229075,
+        "number": 82,
+        "precision": 0.5241379310344828,
+        "recall": 0.926829268292683
+      },
+      "eval_UNIT": {
+        "f1": 0.9566317921025986,
+        "number": 5755,
+        "precision": 0.9298015417418403,
+        "recall": 0.9850564726324935
+      },
+      "eval_loss": 0.48777666687965393,
+      "eval_overall_accuracy": 0.8335786797178198,
+      "eval_overall_f1": 0.8552831783601016,
+      "eval_overall_precision": 0.8288551019070712,
+      "eval_overall_recall": 0.8834520815870355,
+      "eval_runtime": 8.452,
+      "eval_samples_per_second": 1006.385,
+      "eval_steps_per_second": 31.472,
+      "step": 31000
+    },
+    {
+      "epoch": 6.24,
+      "learning_rate": 1.8818055830528607e-05,
+      "loss": 0.4237,
+      "step": 31500
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 1.8323104335775097e-05,
+      "loss": 0.4305,
+      "step": 32000
+    },
+    {
+      "epoch": 6.34,
+      "eval_COMMENT": {
+        "f1": 0.7089205043489988,
+        "number": 6824,
+        "precision": 0.6565505182964906,
+        "recall": 0.770369284876905
+      },
+      "eval_NAME": {
+        "f1": 0.8125769273805529,
+        "number": 8803,
+        "precision": 0.8005732554293904,
+        "recall": 0.8249460411223447
+      },
+      "eval_QTY": {
+        "f1": 0.9859881491808993,
+        "number": 7168,
+        "precision": 0.985369931726348,
+        "recall": 0.9866071428571429
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6634146341463415,
+        "number": 82,
+        "precision": 0.5528455284552846,
+        "recall": 0.8292682926829268
+      },
+      "eval_UNIT": {
+        "f1": 0.956661316211878,
+        "number": 5755,
+        "precision": 0.9309437684972048,
+        "recall": 0.9838401390095569
+      },
+      "eval_loss": 0.4837185740470886,
+      "eval_overall_accuracy": 0.834209570420785,
+      "eval_overall_f1": 0.8570026399512625,
+      "eval_overall_precision": 0.8312869336835194,
+      "eval_overall_recall": 0.8843601564682873,
+      "eval_runtime": 9.3183,
+      "eval_samples_per_second": 912.826,
+      "eval_steps_per_second": 28.546,
+      "step": 32000
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 1.782815284102158e-05,
+      "loss": 0.423,
+      "step": 32500
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 1.7333201346268066e-05,
+      "loss": 0.4142,
+      "step": 33000
+    },
+    {
+      "epoch": 6.53,
+      "eval_COMMENT": {
+        "f1": 0.6982502247734974,
+        "number": 6824,
+        "precision": 0.6611656843483955,
+        "recall": 0.7397420867526378
+      },
+      "eval_NAME": {
+        "f1": 0.8101717599821548,
+        "number": 8803,
+        "precision": 0.7957059918939643,
+        "recall": 0.8251732363966829
+      },
+      "eval_QTY": {
+        "f1": 0.9850003456141564,
+        "number": 7168,
+        "precision": 0.9761611179613646,
+        "recall": 0.9940011160714286
+      },
+      "eval_RANGE_END": {
+        "f1": 0.669683257918552,
+        "number": 82,
+        "precision": 0.5323741007194245,
+        "recall": 0.9024390243902439
+      },
+      "eval_UNIT": {
+        "f1": 0.9568381855111713,
+        "number": 5755,
+        "precision": 0.9326843755155915,
+        "recall": 0.9822762814943528
+      },
+      "eval_loss": 0.48620671033859253,
+      "eval_overall_accuracy": 0.833158085915843,
+      "eval_overall_f1": 0.8545377366499703,
+      "eval_overall_precision": 0.831510425271784,
+      "eval_overall_recall": 0.8788767812238055,
+      "eval_runtime": 6.5738,
+      "eval_samples_per_second": 1293.917,
+      "eval_steps_per_second": 40.463,
+      "step": 33000
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 1.6838249851514553e-05,
+      "loss": 0.4247,
+      "step": 33500
+    },
+    {
+      "epoch": 6.73,
+      "learning_rate": 1.634329835676104e-05,
+      "loss": 0.4327,
+      "step": 34000
+    },
+    {
+      "epoch": 6.73,
+      "eval_COMMENT": {
+        "f1": 0.7027439024390244,
+        "number": 6824,
+        "precision": 0.6665352260778128,
+        "recall": 0.7431125439624854
+      },
+      "eval_NAME": {
+        "f1": 0.8050692012673003,
+        "number": 8803,
+        "precision": 0.7882020026121027,
+        "recall": 0.8226740883789617
+      },
+      "eval_QTY": {
+        "f1": 0.9866000138859959,
+        "number": 7168,
+        "precision": 0.9820317899101589,
+        "recall": 0.9912109375
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6930693069306931,
+        "number": 82,
+        "precision": 0.5833333333333334,
+        "recall": 0.8536585365853658
+      },
+      "eval_UNIT": {
+        "f1": 0.9562352145995269,
+        "number": 5755,
+        "precision": 0.9306035191580332,
+        "recall": 0.9833188531711555
+      },
+      "eval_loss": 0.4834836721420288,
+      "eval_overall_accuracy": 0.8330242606152141,
+      "eval_overall_f1": 0.8544101658059254,
+      "eval_overall_precision": 0.8318007409367557,
+      "eval_overall_recall": 0.8782830399552948,
+      "eval_runtime": 8.5374,
+      "eval_samples_per_second": 996.324,
+      "eval_steps_per_second": 31.157,
+      "step": 34000
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 1.5848346862007525e-05,
+      "loss": 0.4289,
+      "step": 34500
+    },
+    {
+      "epoch": 6.93,
+      "learning_rate": 1.535339536725401e-05,
+      "loss": 0.4299,
+      "step": 35000
+    },
+    {
+      "epoch": 6.93,
+      "eval_COMMENT": {
+        "f1": 0.706956462492335,
+        "number": 6824,
+        "precision": 0.660639246147969,
+        "recall": 0.7602579132473622
+      },
+      "eval_NAME": {
+        "f1": 0.8096777790166128,
+        "number": 8803,
+        "precision": 0.7949644225506295,
+        "recall": 0.8249460411223447
+      },
+      "eval_QTY": {
+        "f1": 0.9858723641171967,
+        "number": 7168,
+        "precision": 0.9836133870295792,
+        "recall": 0.9881417410714286
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6757990867579909,
+        "number": 82,
+        "precision": 0.5401459854014599,
+        "recall": 0.9024390243902439
+      },
+      "eval_UNIT": {
+        "f1": 0.9569053274516457,
+        "number": 5755,
+        "precision": 0.9348582794629537,
+        "recall": 0.9800173761946134
+      },
+      "eval_loss": 0.4832661747932434,
+      "eval_overall_accuracy": 0.8333301470166516,
+      "eval_overall_f1": 0.8559610788086318,
+      "eval_overall_precision": 0.8316150070819197,
+      "eval_overall_recall": 0.8817756356524169,
+      "eval_runtime": 7.9218,
+      "eval_samples_per_second": 1073.752,
+      "eval_steps_per_second": 33.578,
+      "step": 35000
+    },
+    {
+      "epoch": 7.03,
+      "learning_rate": 1.4858443872500496e-05,
+      "loss": 0.4181,
+      "step": 35500
+    },
+    {
+      "epoch": 7.13,
+      "learning_rate": 1.4363492377746981e-05,
+      "loss": 0.4165,
+      "step": 36000
+    },
+    {
+      "epoch": 7.13,
+      "eval_COMMENT": {
+        "f1": 0.7096729965037363,
+        "number": 6824,
+        "precision": 0.6667525441195414,
+        "recall": 0.7584994138335287
+      },
+      "eval_NAME": {
+        "f1": 0.8082838265504414,
+        "number": 8803,
+        "precision": 0.7905082536924414,
+        "recall": 0.8268772009542201
+      },
+      "eval_QTY": {
+        "f1": 0.9864479810966712,
+        "number": 7168,
+        "precision": 0.9828278631768453,
+        "recall": 0.9900948660714286
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6542056074766355,
+        "number": 82,
+        "precision": 0.5303030303030303,
+        "recall": 0.8536585365853658
+      },
+      "eval_UNIT": {
+        "f1": 0.9563968227142133,
+        "number": 5755,
+        "precision": 0.930909689093601,
+        "recall": 0.9833188531711555
+      },
+      "eval_loss": 0.48711806535720825,
+      "eval_overall_accuracy": 0.8347831074234806,
+      "eval_overall_f1": 0.8564749724739562,
+      "eval_overall_precision": 0.8315297832450745,
+      "eval_overall_recall": 0.8829631181894384,
+      "eval_runtime": 8.4247,
+      "eval_samples_per_second": 1009.646,
+      "eval_steps_per_second": 31.574,
+      "step": 36000
+    },
+    {
+      "epoch": 7.23,
+      "learning_rate": 1.3868540882993467e-05,
+      "loss": 0.4199,
+      "step": 36500
+    },
+    {
+      "epoch": 7.33,
+      "learning_rate": 1.3373589388239954e-05,
+      "loss": 0.4213,
+      "step": 37000
+    },
+    {
+      "epoch": 7.33,
+      "eval_COMMENT": {
+        "f1": 0.7081497797356828,
+        "number": 6824,
+        "precision": 0.667705088265836,
+        "recall": 0.753810082063306
+      },
+      "eval_NAME": {
+        "f1": 0.8101195440644984,
+        "number": 8803,
+        "precision": 0.7934001306904814,
+        "recall": 0.827558786777235
+      },
+      "eval_QTY": {
+        "f1": 0.9860821155184412,
+        "number": 7168,
+        "precision": 0.983754512635379,
+        "recall": 0.9884207589285714
+      },
+      "eval_RANGE_END": {
+        "f1": 0.663594470046083,
+        "number": 82,
+        "precision": 0.5333333333333333,
+        "recall": 0.8780487804878049
+      },
+      "eval_UNIT": {
+        "f1": 0.9559865092748736,
+        "number": 5755,
+        "precision": 0.9285831285831286,
+        "recall": 0.9850564726324935
+      },
+      "eval_loss": 0.4881742596626282,
+      "eval_overall_accuracy": 0.8332727933163822,
+      "eval_overall_f1": 0.8566824966078698,
+      "eval_overall_precision": 0.832728831442891,
+      "eval_overall_recall": 0.8820550433081866,
+      "eval_runtime": 8.1883,
+      "eval_samples_per_second": 1038.805,
+      "eval_steps_per_second": 32.486,
+      "step": 37000
+    },
+    {
+      "epoch": 7.42,
+      "learning_rate": 1.287863789348644e-05,
+      "loss": 0.4196,
+      "step": 37500
+    },
+    {
+      "epoch": 7.52,
+      "learning_rate": 1.2383686398732925e-05,
+      "loss": 0.4122,
+      "step": 38000
+    },
+    {
+      "epoch": 7.52,
+      "eval_COMMENT": {
+        "f1": 0.7068037866920929,
+        "number": 6824,
+        "precision": 0.660262119862578,
+        "recall": 0.7604044548651817
+      },
+      "eval_NAME": {
+        "f1": 0.8086100450525612,
+        "number": 8803,
+        "precision": 0.7921752397558849,
+        "recall": 0.8257412245825286
+      },
+      "eval_QTY": {
+        "f1": 0.9867148918411351,
+        "number": 7168,
+        "precision": 0.9839090026355944,
+        "recall": 0.9895368303571429
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6766169154228855,
+        "number": 82,
+        "precision": 0.5714285714285714,
+        "recall": 0.8292682926829268
+      },
+      "eval_UNIT": {
+        "f1": 0.9566393373341222,
+        "number": 5755,
+        "precision": 0.9313693219223174,
+        "recall": 0.9833188531711555
+      },
+      "eval_loss": 0.4913991689682007,
+      "eval_overall_accuracy": 0.8336933871183589,
+      "eval_overall_f1": 0.8558514330212795,
+      "eval_overall_precision": 0.8304477808075167,
+      "eval_overall_recall": 0.8828583403185247,
+      "eval_runtime": 6.6374,
+      "eval_samples_per_second": 1281.524,
+      "eval_steps_per_second": 40.076,
+      "step": 38000
+    },
+    {
+      "epoch": 7.62,
+      "learning_rate": 1.1888734903979411e-05,
+      "loss": 0.4196,
+      "step": 38500
+    },
+    {
+      "epoch": 7.72,
+      "learning_rate": 1.1393783409225896e-05,
+      "loss": 0.4034,
+      "step": 39000
+    },
+    {
+      "epoch": 7.72,
+      "eval_COMMENT": {
+        "f1": 0.7121398334016114,
+        "number": 6824,
+        "precision": 0.6667092815136794,
+        "recall": 0.7642145369284877
+      },
+      "eval_NAME": {
+        "f1": 0.8117607676857844,
+        "number": 8803,
+        "precision": 0.7976099111939481,
+        "recall": 0.8264228104055436
+      },
+      "eval_QTY": {
+        "f1": 0.9866388308977034,
+        "number": 7168,
+        "precision": 0.984309913912802,
+        "recall": 0.9889787946428571
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6859903381642511,
+        "number": 82,
+        "precision": 0.568,
+        "recall": 0.8658536585365854
+      },
+      "eval_UNIT": {
+        "f1": 0.9563307711799983,
+        "number": 5755,
+        "precision": 0.9304733727810651,
+        "recall": 0.9836663770634231
+      },
+      "eval_loss": 0.4883587956428528,
+      "eval_overall_accuracy": 0.8344007494216835,
+      "eval_overall_f1": 0.8582036415420609,
+      "eval_overall_precision": 0.8338604467286025,
+      "eval_overall_recall": 0.884010896898575,
+      "eval_runtime": 9.4932,
+      "eval_samples_per_second": 896.008,
+      "eval_steps_per_second": 28.02,
+      "step": 39000
+    },
+    {
+      "epoch": 7.82,
+      "learning_rate": 1.0898831914472382e-05,
+      "loss": 0.4145,
+      "step": 39500
+    },
+    {
+      "epoch": 7.92,
+      "learning_rate": 1.0403880419718868e-05,
+      "loss": 0.4209,
+      "step": 40000
+    },
+    {
+      "epoch": 7.92,
+      "eval_COMMENT": {
+        "f1": 0.706067877956805,
+        "number": 6824,
+        "precision": 0.6634454322896534,
+        "recall": 0.7545427901524033
+      },
+      "eval_NAME": {
+        "f1": 0.81135449262405,
+        "number": 8803,
+        "precision": 0.7984163642362256,
+        "recall": 0.8247188458480064
+      },
+      "eval_QTY": {
+        "f1": 0.9865127919911012,
+        "number": 7168,
+        "precision": 0.9832317073170732,
+        "recall": 0.9898158482142857
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6697674418604651,
+        "number": 82,
+        "precision": 0.5413533834586466,
+        "recall": 0.8780487804878049
+      },
+      "eval_UNIT": {
+        "f1": 0.9571609632446134,
+        "number": 5755,
+        "precision": 0.9315789473684211,
+        "recall": 0.9841876629018245
+      },
+      "eval_loss": 0.49103492498397827,
+      "eval_overall_accuracy": 0.8326992563136865,
+      "eval_overall_f1": 0.8568276330306375,
+      "eval_overall_precision": 0.8334709242809497,
+      "eval_overall_recall": 0.8815311539536184,
+      "eval_runtime": 6.7729,
+      "eval_samples_per_second": 1255.88,
+      "eval_steps_per_second": 39.274,
+      "step": 40000
+    },
+    {
+      "epoch": 8.02,
+      "learning_rate": 9.908928924965353e-06,
+      "loss": 0.4146,
+      "step": 40500
+    },
+    {
+      "epoch": 8.12,
+      "learning_rate": 9.41397743021184e-06,
+      "loss": 0.4078,
+      "step": 41000
+    },
+    {
+      "epoch": 8.12,
+      "eval_COMMENT": {
+        "f1": 0.7065016749846176,
+        "number": 6824,
+        "precision": 0.6621812123542228,
+        "recall": 0.7571805392731535
+      },
+      "eval_NAME": {
+        "f1": 0.8095132004010249,
+        "number": 8803,
+        "precision": 0.7941208611080757,
+        "recall": 0.8255140293081904
+      },
+      "eval_QTY": {
+        "f1": 0.9861565217391305,
+        "number": 7168,
+        "precision": 0.9834882752879145,
+        "recall": 0.9888392857142857
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6824644549763034,
+        "number": 82,
+        "precision": 0.5581395348837209,
+        "recall": 0.8780487804878049
+      },
+      "eval_UNIT": {
+        "f1": 0.9566246723598546,
+        "number": 5755,
+        "precision": 0.9316534914361001,
+        "recall": 0.982971329278888
+      },
+      "eval_loss": 0.49227482080459595,
+      "eval_overall_accuracy": 0.8334830902173705,
+      "eval_overall_f1": 0.856053157948266,
+      "eval_overall_precision": 0.8316645807259074,
+      "eval_overall_recall": 0.8819153394803018,
+      "eval_runtime": 8.7342,
+      "eval_samples_per_second": 973.868,
+      "eval_steps_per_second": 30.455,
+      "step": 41000
+    },
+    {
+      "epoch": 8.22,
+      "learning_rate": 8.919025935458326e-06,
+      "loss": 0.4008,
+      "step": 41500
+    },
+    {
+      "epoch": 8.32,
+      "learning_rate": 8.424074440704812e-06,
+      "loss": 0.4098,
+      "step": 42000
+    },
+    {
+      "epoch": 8.32,
+      "eval_COMMENT": {
+        "f1": 0.7069577080491133,
+        "number": 6824,
+        "precision": 0.6613067891781521,
+        "recall": 0.7593786635404455
+      },
+      "eval_NAME": {
+        "f1": 0.8097364895042429,
+        "number": 8803,
+        "precision": 0.7961356899769458,
+        "recall": 0.8238100647506532
+      },
+      "eval_QTY": {
+        "f1": 0.9860801781737194,
+        "number": 7168,
+        "precision": 0.9838888888888889,
+        "recall": 0.98828125
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6889952153110047,
+        "number": 82,
+        "precision": 0.5669291338582677,
+        "recall": 0.8780487804878049
+      },
+      "eval_UNIT": {
+        "f1": 0.9567859554355165,
+        "number": 5755,
+        "precision": 0.9302478253733792,
+        "recall": 0.9848827106863597
+      },
+      "eval_loss": 0.4903165102005005,
+      "eval_overall_accuracy": 0.8344198673217734,
+      "eval_overall_f1": 0.8562469278098886,
+      "eval_overall_precision": 0.831812942532521,
+      "eval_overall_recall": 0.8821598211791003,
+      "eval_runtime": 7.5483,
+      "eval_samples_per_second": 1126.874,
+      "eval_steps_per_second": 35.24,
+      "step": 42000
+    },
+    {
+      "epoch": 8.41,
+      "learning_rate": 7.929122945951298e-06,
+      "loss": 0.4092,
+      "step": 42500
+    },
+    {
+      "epoch": 8.51,
+      "learning_rate": 7.434171451197783e-06,
+      "loss": 0.4009,
+      "step": 43000
+    },
+    {
+      "epoch": 8.51,
+      "eval_COMMENT": {
+        "f1": 0.7111504907306434,
+        "number": 6824,
+        "precision": 0.6647553516819572,
+        "recall": 0.7645076201641267
+      },
+      "eval_NAME": {
+        "f1": 0.8115311698449873,
+        "number": 8803,
+        "precision": 0.7969554265688314,
+        "recall": 0.8266500056798819
+      },
+      "eval_QTY": {
+        "f1": 0.9865015307542444,
+        "number": 7168,
+        "precision": 0.9840366463076069,
+        "recall": 0.9889787946428571
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6666666666666667,
+        "number": 82,
+        "precision": 0.5689655172413793,
+        "recall": 0.8048780487804879
+      },
+      "eval_UNIT": {
+        "f1": 0.9566759564225994,
+        "number": 5755,
+        "precision": 0.9306605323693723,
+        "recall": 0.9841876629018245
+      },
+      "eval_loss": 0.4936346113681793,
+      "eval_overall_accuracy": 0.8347639895233907,
+      "eval_overall_f1": 0.8578206279546571,
+      "eval_overall_precision": 0.8330755306894849,
+      "eval_overall_recall": 0.8840807488125174,
+      "eval_runtime": 7.8937,
+      "eval_samples_per_second": 1077.574,
+      "eval_steps_per_second": 33.698,
+      "step": 43000
+    },
+    {
+      "epoch": 8.61,
+      "learning_rate": 6.939219956444269e-06,
+      "loss": 0.4103,
+      "step": 43500
+    },
+    {
+      "epoch": 8.71,
+      "learning_rate": 6.444268461690754e-06,
+      "loss": 0.41,
+      "step": 44000
+    },
+    {
+      "epoch": 8.71,
+      "eval_COMMENT": {
+        "f1": 0.7139846352573254,
+        "number": 6824,
+        "precision": 0.6659480025364616,
+        "recall": 0.7694900351699883
+      },
+      "eval_NAME": {
+        "f1": 0.8144254962258876,
+        "number": 8803,
+        "precision": 0.8019158775600088,
+        "recall": 0.8273315915028967
+      },
+      "eval_QTY": {
+        "f1": 0.9864385562278323,
+        "number": 7168,
+        "precision": 0.9834974344751075,
+        "recall": 0.9893973214285714
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6728110599078342,
+        "number": 82,
+        "precision": 0.5407407407407407,
+        "recall": 0.8902439024390244
+      },
+      "eval_UNIT": {
+        "f1": 0.9569402228976698,
+        "number": 5755,
+        "precision": 0.9306946953522746,
+        "recall": 0.9847089487402259
+      },
+      "eval_loss": 0.4894082546234131,
+      "eval_overall_accuracy": 0.835911063528782,
+      "eval_overall_f1": 0.8593691770843921,
+      "eval_overall_precision": 0.8343530030919019,
+      "eval_overall_recall": 0.8859318245319922,
+      "eval_runtime": 9.03,
+      "eval_samples_per_second": 941.975,
+      "eval_steps_per_second": 29.457,
+      "step": 44000
+    },
+    {
+      "epoch": 8.81,
+      "learning_rate": 5.949316966937241e-06,
+      "loss": 0.4063,
+      "step": 44500
+    },
+    {
+      "epoch": 8.91,
+      "learning_rate": 5.4543654721837265e-06,
+      "loss": 0.4177,
+      "step": 45000
+    },
+    {
+      "epoch": 8.91,
+      "eval_COMMENT": {
+        "f1": 0.7089166264395559,
+        "number": 6824,
+        "precision": 0.6695323694151362,
+        "recall": 0.7532239155920282
+      },
+      "eval_NAME": {
+        "f1": 0.810614930032893,
+        "number": 8803,
+        "precision": 0.7959273045763083,
+        "recall": 0.8258548222196979
+      },
+      "eval_QTY": {
+        "f1": 0.9862956521739129,
+        "number": 7168,
+        "precision": 0.9836270292770917,
+        "recall": 0.9889787946428571
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6926829268292684,
+        "number": 82,
+        "precision": 0.5772357723577236,
+        "recall": 0.8658536585365854
+      },
+      "eval_UNIT": {
+        "f1": 0.9571682014023823,
+        "number": 5755,
+        "precision": 0.9314370272936534,
+        "recall": 0.9843614248479583
+      },
+      "eval_loss": 0.4893593490123749,
+      "eval_overall_accuracy": 0.8339036840193473,
+      "eval_overall_f1": 0.8575312207968737,
+      "eval_overall_precision": 0.8349601297025444,
+      "eval_overall_recall": 0.8813565241687622,
+      "eval_runtime": 7.9862,
+      "eval_samples_per_second": 1065.084,
+      "eval_steps_per_second": 33.307,
+      "step": 45000
+    },
+    {
+      "epoch": 9.01,
+      "learning_rate": 4.959413977430212e-06,
+      "loss": 0.4109,
+      "step": 45500
+    },
+    {
+      "epoch": 9.11,
+      "learning_rate": 4.4644624826766974e-06,
+      "loss": 0.399,
+      "step": 46000
+    },
+    {
+      "epoch": 9.11,
+      "eval_COMMENT": {
+        "f1": 0.7139457401237506,
+        "number": 6824,
+        "precision": 0.6659901052898642,
+        "recall": 0.7693434935521688
+      },
+      "eval_NAME": {
+        "f1": 0.8108047897521581,
+        "number": 8803,
+        "precision": 0.7953452797202797,
+        "recall": 0.8268772009542201
+      },
+      "eval_QTY": {
+        "f1": 0.9858156028368794,
+        "number": 7168,
+        "precision": 0.9826725810923205,
+        "recall": 0.9889787946428571
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6790697674418604,
+        "number": 82,
+        "precision": 0.5488721804511278,
+        "recall": 0.8902439024390244
+      },
+      "eval_UNIT": {
+        "f1": 0.9572476600050595,
+        "number": 5755,
+        "precision": 0.9298820445609436,
+        "recall": 0.98627280625543
+      },
+      "eval_loss": 0.49086296558380127,
+      "eval_overall_accuracy": 0.8356434129275241,
+      "eval_overall_f1": 0.8581819412023411,
+      "eval_overall_precision": 0.832086859542085,
+      "eval_overall_recall": 0.8859667504889635,
+      "eval_runtime": 9.6791,
+      "eval_samples_per_second": 878.799,
+      "eval_steps_per_second": 27.482,
+      "step": 46000
+    },
+    {
+      "epoch": 9.21,
+      "learning_rate": 3.969510987923184e-06,
+      "loss": 0.4028,
+      "step": 46500
+    },
+    {
+      "epoch": 9.31,
+      "learning_rate": 3.4745594931696697e-06,
+      "loss": 0.4066,
+      "step": 47000
+    },
+    {
+      "epoch": 9.31,
+      "eval_COMMENT": {
+        "f1": 0.7096332785987959,
+        "number": 6824,
+        "precision": 0.6655544147843943,
+        "recall": 0.7599648300117233
+      },
+      "eval_NAME": {
+        "f1": 0.8115457483561797,
+        "number": 8803,
+        "precision": 0.7964563053702286,
+        "recall": 0.8272179938657276
+      },
+      "eval_QTY": {
+        "f1": 0.9858900396191006,
+        "number": 7168,
+        "precision": 0.9824075356697604,
+        "recall": 0.9893973214285714
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6792452830188679,
+        "number": 82,
+        "precision": 0.5538461538461539,
+        "recall": 0.8780487804878049
+      },
+      "eval_UNIT": {
+        "f1": 0.9569838056680162,
+        "number": 5755,
+        "precision": 0.9298475659727914,
+        "recall": 0.9857515204170286
+      },
+      "eval_loss": 0.4892035126686096,
+      "eval_overall_accuracy": 0.8339610377196168,
+      "eval_overall_f1": 0.8575495196299372,
+      "eval_overall_precision": 0.8328122428830015,
+      "eval_overall_recall": 0.8838013411567477,
+      "eval_runtime": 7.0853,
+      "eval_samples_per_second": 1200.507,
+      "eval_steps_per_second": 37.542,
+      "step": 47000
+    },
+    {
+      "epoch": 9.4,
+      "learning_rate": 2.979607998416155e-06,
+      "loss": 0.4045,
+      "step": 47500
+    },
+    {
+      "epoch": 9.5,
+      "learning_rate": 2.484656503662641e-06,
+      "loss": 0.397,
+      "step": 48000
+    },
+    {
+      "epoch": 9.5,
+      "eval_COMMENT": {
+        "f1": 0.7098901098901098,
+        "number": 6824,
+        "precision": 0.6680455015511892,
+        "recall": 0.757327080890973
+      },
+      "eval_NAME": {
+        "f1": 0.8101547711836098,
+        "number": 8803,
+        "precision": 0.7944098700731521,
+        "recall": 0.8265364080427127
+      },
+      "eval_QTY": {
+        "f1": 0.9862327909887358,
+        "number": 7168,
+        "precision": 0.9830884391461048,
+        "recall": 0.9893973214285714
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6857142857142857,
+        "number": 82,
+        "precision": 0.5625,
+        "recall": 0.8780487804878049
+      },
+      "eval_UNIT": {
+        "f1": 0.9567713610266801,
+        "number": 5755,
+        "precision": 0.9305304647725406,
+        "recall": 0.9845351867940921
+      },
+      "eval_loss": 0.491767019033432,
+      "eval_overall_accuracy": 0.8346875179230313,
+      "eval_overall_f1": 0.8573560839919944,
+      "eval_overall_precision": 0.8334102750115413,
+      "eval_overall_recall": 0.8827186364906399,
+      "eval_runtime": 8.8099,
+      "eval_samples_per_second": 965.502,
+      "eval_steps_per_second": 30.193,
+      "step": 48000
+    },
+    {
+      "epoch": 9.6,
+      "learning_rate": 1.989705008909127e-06,
+      "loss": 0.4023,
+      "step": 48500
+    },
+    {
+      "epoch": 9.7,
+      "learning_rate": 1.4947535141556129e-06,
+      "loss": 0.3987,
+      "step": 49000
+    },
+    {
+      "epoch": 9.7,
+      "eval_COMMENT": {
+        "f1": 0.7078929306794783,
+        "number": 6824,
+        "precision": 0.6657629744384198,
+        "recall": 0.7557151230949589
+      },
+      "eval_NAME": {
+        "f1": 0.8108077994428969,
+        "number": 8803,
+        "precision": 0.7955613862468569,
+        "recall": 0.8266500056798819
+      },
+      "eval_QTY": {
+        "f1": 0.9860232250886588,
+        "number": 7168,
+        "precision": 0.9829474559822543,
+        "recall": 0.9891183035714286
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6857142857142857,
+        "number": 82,
+        "precision": 0.5625,
+        "recall": 0.8780487804878049
+      },
+      "eval_UNIT": {
+        "f1": 0.9566759564225994,
+        "number": 5755,
+        "precision": 0.9306605323693723,
+        "recall": 0.9841876629018245
+      },
+      "eval_loss": 0.4935953915119171,
+      "eval_overall_accuracy": 0.8338845661192574,
+      "eval_overall_f1": 0.8569683810557741,
+      "eval_overall_precision": 0.83311345646438,
+      "eval_overall_recall": 0.8822296730930428,
+      "eval_runtime": 7.6552,
+      "eval_samples_per_second": 1111.139,
+      "eval_steps_per_second": 34.748,
+      "step": 49000
+    },
+    {
+      "epoch": 9.8,
+      "learning_rate": 9.998020194020988e-07,
+      "loss": 0.3986,
+      "step": 49500
+    },
+    {
+      "epoch": 9.9,
+      "learning_rate": 5.048505246485845e-07,
+      "loss": 0.4109,
+      "step": 50000
+    },
+    {
+      "epoch": 9.9,
+      "eval_COMMENT": {
+        "f1": 0.7082043875937006,
+        "number": 6824,
+        "precision": 0.6672281974860698,
+        "recall": 0.7545427901524033
+      },
+      "eval_NAME": {
+        "f1": 0.8094707520891365,
+        "number": 8803,
+        "precision": 0.794249480704056,
+        "recall": 0.8252868340338521
+      },
+      "eval_QTY": {
+        "f1": 0.986230876216968,
+        "number": 7168,
+        "precision": 0.983222407099279,
+        "recall": 0.9892578125
+      },
+      "eval_RANGE_END": {
+        "f1": 0.6889952153110047,
+        "number": 82,
+        "precision": 0.5669291338582677,
+        "recall": 0.8780487804878049
+      },
+      "eval_UNIT": {
+        "f1": 0.956595169734842,
+        "number": 5755,
+        "precision": 0.9305076392311483,
+        "recall": 0.9841876629018245
+      },
+      "eval_loss": 0.49291422963142395,
+      "eval_overall_accuracy": 0.8339228019194371,
+      "eval_overall_f1": 0.8567597841213808,
+      "eval_overall_precision": 0.8333113238692638,
+      "eval_overall_recall": 0.8815660799105896,
+      "eval_runtime": 7.4834,
+      "eval_samples_per_second": 1136.652,
+      "eval_steps_per_second": 35.545,
+      "step": 50000
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 9.899029895070283e-09,
+      "loss": 0.4034,
+      "step": 50500
+    },
+    {
+      "epoch": 10.0,
+      "step": 50510,
+      "total_flos": 1557854363345160.0,
+      "train_loss": 0.46329507851359913,
+      "train_runtime": 1951.481,
+      "train_samples_per_second": 828.104,
+      "train_steps_per_second": 25.883
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 50510,
+  "num_train_epochs": 10,
+  "save_steps": 1000,
+  "total_flos": 1557854363345160.0,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1b1785c20e6b7459a42abe5f3a6eeb47964ef4f0d5ec55128274078021755b9c
+size 4536

validation_results.json

@@ -0,0 +1,41 @@
+{
+    "epoch": 10.0,
+    "eval_COMMENT": {
+        "f1": 0.7049292873815736,
+        "number": 6824,
+        "precision": 0.6631361405321622,
+        "recall": 0.7523446658851114
+    },
+    "eval_NAME": {
+        "f1": 0.8063276332646355,
+        "number": 8803,
+        "precision": 0.7910382513661203,
+        "recall": 0.8222196978302851
+    },
+    "eval_QTY": {
+        "f1": 0.9859881491808993,
+        "number": 7168,
+        "precision": 0.985369931726348,
+        "recall": 0.9866071428571429
+    },
+    "eval_RANGE_END": {
+        "f1": 0.6783625730994152,
+        "number": 82,
+        "precision": 0.651685393258427,
+        "recall": 0.7073170731707317
+    },
+    "eval_UNIT": {
+        "f1": 0.9558001693480102,
+        "number": 5755,
+        "precision": 0.9321222130470685,
+        "recall": 0.9807124239791486
+    },
+    "eval_loss": 0.47966957092285156,
+    "eval_overall_accuracy": 0.832374252012159,
+    "eval_overall_f1": 0.8546520520010196,
+    "eval_overall_precision": 0.8322907357759904,
+    "eval_overall_recall": 0.8782481139983236,
+    "eval_runtime": 7.9043,
+    "eval_samples_per_second": 1076.13,
+    "eval_steps_per_second": 33.653
+}