115 lines
18 KiB
Plaintext
115 lines
18 KiB
Plaintext
|
Ë
|
|||
|
|
?û g7< ã ó¤ — d dl Z d dlZd dlZd dlmZ d dlmZ d dlm Z m
|
|||
|
|
Z
|
|||
|
|
m Z m Z m
|
|||
|
|
Z
|
|||
|
|
mZmZmZmZmZ d dlmZmZmZmZmZ d dlmZ ddlm Z dd l m Z dd
|
|||
|
|
l m Z m!Z!m"Z" d d l#m$Z$ e
|
|||
|
|
rdd
|
|||
|
|
l%m&Z& dZ'dZ( d3dejR ejT dœdddee de+de de dedee f f d„Z,dddede-d e+d e+d e+d e+d ee. d!ee. d"ee de
|
|||
|
|
e.e f gdf fd#„Z/d$„ Z0ddd%e d&e
|
|||
|
|
e.e-f de g ee-e
|
|||
|
|
e.e-f f f fd'„Z1ddd(e dgee$ f d)e ee$ gee$ f d*e+fd+„Z2d,ee
|
|||
|
|
e.e f ef ddd-e
|
|||
|
|
e.e f ddfd.„Z3d/ee dgdf de dgdf fd0„Z4d1ee ddfd2„Z5y)4é N)ÚPath)Ú
|
|||
|
|
default_timer)
|
|||
|
|
ÚIOÚ
|
|||
|
|
TYPE_CHECKINGÚAnyÚCallableÚDictÚIterableÚListÚOptionalÚTupleÚUnion)ÚConfigÚ OptimizerÚconstantÚfix_random_seedÚset_gpu_allocator)ÚPrinteré )ÚErrors)ÚConfigSchemaTraining)ÚloggerÚregistryÚresolve_dot_namesé )ÚExample©ÚLanguagez
|
|||
|
|
model-bestz
|
|||
|
|
model-lastéÿÿÿÿ)Úuse_gpuÚstdoutÚstderrÚnlpr Ú output_pathr r! r" Úreturnc óv ‡ ‡‡‡— t d¬« }‰ j j « }|d d �t |d d « |d d }|dk\ r
|
|||
|
|
|r t |« t j |d t ¬« }|d |d
|
|||
|
|
g} t || « \ }
|
|||
|
|
} |d Š|d } |d
|
|||
|
|
}
|
|||
|
|
|d }t |d « Š|d }ˆˆ ˆˆfd„}|d }|d }t ‰ ‰t ‰ |
|
|||
|
|
|
|
|||
|
|
|d « t ‰ | | « |d |d |d |d |d |||¬« }t ‰« |j |j d‰ j › �« d z « |r&|j |j d |› �« d z « |r&|j |j d |› �« d z « |j |j d ‰j" › �« d z « ‰ j% |¬ « 5 |‰ ||« \ }}ddd« |D ][ \ }}}|�G‰ j% |¬ « 5 t' |‰ |« ddd« ‰� ||« t) ‰t* z « |d!< |�|nd« Œ] « ‰� |d#« ‰j2 r‰ j5 ‰j2 « ‰�6|j |j7 d$‰t* z « d z « ‰ ‰t* z fS ‰ dfS # 1 sw Y ŒàxY w# 1 sw Y Œ¸xY w# t, $ r8}‰�/|j |j/ d"t1 |« › �« d z « |‚d}~ww xY w# « ‰� |d#« w w xY w)%aU Train a pipeline.
|
|||
|
|
|
|||
|
|
nlp (Language): The initialized nlp object with the full config.
|
|||
|
|
output_path (Optional[Path]): Optional output path to save trained model to.
|
|||
|
|
use_gpu (int): Whether to train on GPU. Make sure to call require_gpu
|
|||
|
|
before calling this function.
|
|||
|
|
stdout (file): A file-like object to write output messages. To disable
|
|||
|
|
printing, set to io.StringIO.
|
|||
|
|
stderr (file): A second file-like object to write output messages. To disable
|
|||
|
|
printing, set to io.StringIO.
|
|||
|
|
|
|||
|
|
RETURNS (tuple): The final nlp object and the path to the exported model.
|
|||
|
|
T)Úno_printÚtrainingÚseedNÚ
|
|||
|
|
gpu_allocatorr )ÚschemaÚ train_corpusÚ
|
|||
|
|
dev_corpusÚ optimizerÚ
|
|||
|
|
score_weightsÚbatcherr Úbefore_to_diskÚ
|
|||
|
|
before_updatec óV •— ‰j ‰j « 5 ‰‰« j ‰t z « d d d « | rX‰t z j « r t
|
|||
|
|
j ‰t z « t
|
|||
|
|
j ‰t z ‰t z « y y # 1 sw Y ŒdxY w©N) Ú
|
|||
|
|
use_paramsÚaveragesÚto_diskÚDIR_MODEL_LASTÚDIR_MODEL_BESTÚexistsÚshutilÚrmtreeÚcopytree)Úis_bestr1 r# r. r$ s €€€€úTC:\Users\garci\AppData\Roaming\Python\Python312\site-packages\spacy/training/loop.pyÚsave_checkpointz train.<locals>.save_checkpointM s ø€ Ø
|
|||
|
|
�^‰^˜I× .Ñ .Õ
|
|||
|
|
/Ù ˜3Ó × 'Ñ '¨ ´nÑ(DÔ E÷ 0á ð œnÑ,×4Ñ4Ô6Ü—
|
|||
|
|
‘
|
|||
|
|
˜k¬NÑ :Ô;Ü �O‰O˜K¬.Ñ 8¸+ÌÑ:VÕ Wð ÷ 0Ð
|
|||
|
|
/ús � B Â B(Úfrozen_componentsÚannotating_componentsÚ
|
|||
|
|
max_epochsÚdropoutÚaccumulate_gradientÚpatienceÚ max_stepsÚeval_frequency)rD rE rF rG rH ÚexcluderB r2 z
|
|||
|
|
Pipeline: ú
|
|||
|
|
zFrozen components: z Set annotations on update for: zInitial learn rate: )Údisabler$ zAAborting and saving the final best model. Encountered exception: Fz"Saved pipeline to output directory) r ÚconfigÚ interpolater r r Úresolver r Ú create_before_to_disk_callbackÚtrain_while_improvingÚcreate_train_batchesÚcreate_evaluation_callbackÚclean_output_dirÚwriteÚinfoÚ
|
|||
|
|
pipe_namesÚ
|
|||
|
|
learn_rateÚ select_pipesÚ update_metaÚstrr8 Ú ExceptionÚwarnÚreprr6 r5 Úgood) r# r$ r r! r" ÚmsgrL Ú allocatorÚTÚ dot_namesr, r- r/ r0 Ú train_loggerr2 r@ rA rB Útraining_step_iteratorÚlog_stepÚfinalize_loggerÚbatchrU Úis_best_checkpointÚer1 r. s `` @@r? Útrainrj # s§ û€ ô, ˜4Ô
|
|||
|
|
€Cà
|
|||
|
|
�Z‰Z×
|
|||
|
|
#Ñ
|
|||
|
|
#Ó
|
|||
|
|
%€FØ
|
|||
|
|
ˆjј&Ñ!Ð-ܘ˜zÑ*¨6Ñ2Ô3Ø�zÑ" ?Ñ3€IØ�!‚|™ ܘ)Ô$Ü×ј
|
|||
|
|
Ñ+Ô4HÔI€AØ�>Ñ" A l¡OÐ4€IÜ 0°¸Ó CÑ €L�*Ø�+‘€IØ�oÑ&€MØ� ‰l€GØ�X‘;€LÜ3°AÐ6FÑ4GÓH€NØ�oÑ&€M÷Xð Ð-Ñ.Ðà Ð 5Ñ 6Ðä 2Ø ØÜ ˜S ,°¸¸<¹ ÓIÜ" 3¨
|
|||
|
|
°MÓBØ�)‘ Ø Ð 3Ñ 4Ø�:‘ Ø�K‘.ØÐ)Ñ*Ø!Ø 3Ø#ô
|
|||
|
|
Ðô �[Ô!Ø
|
|||
|
|
‡L�L�—‘˜J s§~¡~Ð&6Ð7Ó8¸4Ñ?Ô@ÙØ� ‰ �S—X‘XÐ 3Ð4EÐ3FÐ GÓHÈ4ÑOÔPÙ Ø� ‰ Ø �H‰HÐ6Ð7LÐ6MÐNÓ OÐRVÑ Vô
|
|||
|
|
ð ‡L�L�—‘Ð 0°×1EÑ1EÐ0FÐGÓHÈ4ÑOÔPØ × Ñ Ð"3Ð Õ 4Ù$0°°f¸fÓ$EÑ!ˆ�/÷
|
|||
|
|
5ð#Û/EÑ +ˆE�4Ð+Ø!Ð-Ø×%Ñ%Ð.?Ð%Õ@Ü 3¨Ô-÷ Aà Ð*Ù#Ð$6Ô7Ü*-¨k¼NÑ.JÓ*K�D˜ Ñ'Ù Ð /Ð ;‘TÀÕ Fñ 0Fñ& ÔØ Ð "Ù ˜EÕ "à×ÒØ �‰�y×)Ñ)Ô*ØÐ Ø� ‰ Ø �H‰HÐ9¸;Ì Ñ;WÓ XØñ
|
|||
|
|
ô
|
|||
|
|
ð �[¤>Ñ1Ð2Ð2à�Tˆ{Ð÷G
|
|||
|
|
5Ð 4ú÷
|
|||
|
|
AÐ@ûô ò Ø Ð "Ø �L‰LØ—‘ð.Ü.2°1«g¨Yð8óð ñ ô
|
|||
|
|
ð ˆûð ûñ ÔØ Ð "Ù ˜EÕ "ð #úsH ÇKÇ2 K ÈKÈ 5K ËKËK ËK Ë L!Ë)3L Ì L!Ì!L$ Ì$L8r. rD rH rE rF rG rI rB r2 c # óÄ K — t |t « r t |« } n|} g }
|
|||
|
|
i }d}t « }t |« D �]“ \ }\ }}| r||dœ} | | |« t | « }t
|
|||
|
|
||« D ] }| j |||d| |
|
|||
|
|
¬« Œ | j D ]D \ }}|| vsŒ t |d« sŒ|j sŒ%|j dvsŒ4|j |« ŒF |j « ||z sj|j r/| j |j « 5 |« \ }}ddd« n
|
|||
|
|
|« \ }}|
|
|||
|
|
j! |f« |t# |
|
|||
|
|
« d k( }nd\ }}d}|t% d „ |D « « z
|
|||
|
|
}|||||
|
|||
|
|
t' t « |z
|
|||
|
|
« |d
|
|||
|
|
œ}|||f–— |�i }t# d „ |
|
|||
|
|
D « « } | d } |r
|
|||
|
|
|| z
|
|||
|
|
|k\ r y|s�Œ�||k\ s�Œ” y y# 1 sw Y Œ§xY ww)
|
|||
|
|
a Train until an evaluation stops improving. Works as a generator,
|
|||
|
|
with each iteration yielding a tuple `(batch, info, is_best_checkpoint)`,
|
|||
|
|
where info is a dict, and is_best_checkpoint is in [True, False, None] --
|
|||
|
|
None indicating that the iteration was not evaluated as a checkpoint.
|
|||
|
|
The evaluation is conducted by calling the evaluate callback.
|
|||
|
|
|
|||
|
|
Positional arguments:
|
|||
|
|
nlp: The spaCy pipeline to evaluate.
|
|||
|
|
optimizer: The optimizer callable.
|
|||
|
|
train_data (Iterable[Batch]): A generator of batches, with the training
|
|||
|
|
data. Each batch should be a Sized[Tuple[Input, Annot]]. The training
|
|||
|
|
data iterable needs to take care of iterating over the epochs and
|
|||
|
|
shuffling.
|
|||
|
|
evaluate (Callable[[], Tuple[float, Any]]): A callback to perform evaluation.
|
|||
|
|
The callback should take no arguments and return a tuple
|
|||
|
|
`(main_score, other_scores)`. The main_score should be a float where
|
|||
|
|
higher is better. other_scores can be any object.
|
|||
|
|
|
|||
|
|
Every iteration, the function yields out a tuple with:
|
|||
|
|
|
|||
|
|
* batch: A list of Example objects.
|
|||
|
|
* info: A dict with various information about the last update (see below).
|
|||
|
|
* is_best_checkpoint: A value in None, False, True, indicating whether this
|
|||
|
|
was the best evaluation so far. You should use this to save the model
|
|||
|
|
checkpoints during training. If None, evaluation was not conducted on
|
|||
|
|
that iteration. False means evaluation was conducted, but a previous
|
|||
|
|
evaluation was better.
|
|||
|
|
|
|||
|
|
The info dict provides the following information:
|
|||
|
|
|