+
QV-j@ ^ RI t ^ RIt^ RIt^ RIHtHt ^ RIHt ^RI H
t
^RI H t H
t
]! R4 tR R ltR tRR ltR
t ! R R 4 t ! R
R]4 t ! R R]
4 tR# ) N)OptionalTypeVar)TrainerCallback)IntervalStrategy
has_length_Tc J V ^8 d QhR\ R, R\ R\ /# ) xNmsgreturn)r str)formats "l/Volumes/fast/ai/experiments/ui-tars-smoke/.venv/lib/python3.14/site-packages/transformers/utils/notebook.py __annotate__r s%
T
c $ V f \ V4 hV # N) RuntimeError)r r s &&r _requirer s y3 Hr c \ V 4 p V R, V ^<, ^<, V ^<, r2pV^ 8w d
V RVR
RVR
2# VR
RVR
2# )z$Format `t` (in seconds) to (h):mm:ssi :02d)int)thmss & r format_timer # s[ AA4i!r'R R!A%&!VaS!C!C !CAc7!Ac71CCr c & R V RV RV RV RV R2 # )z
)valuetotalprefixlabelwidths &&&&&r html_progress_barr' * s<
h wgeWOE7 C g r c 4 RpVR,
pV ^ , F pVRV R2,
pK VR,
pV R, FP pVR,
pV F5 p\ V\ 4 ' d VR
M
\ V4 pVR V R
2,
pK7 VR ,
pKR VR ,
pV# )
z*Put the texts in `items` in an HTML table.z%)
isinstancefloatr )items html_codeilineelts & r text_to_html_tabler2 5 s
The main method to update the progress bar to `value`.
Args:
value (`int`):
The value to use. Must be between 0 and `total`.
force_update (`bool`, *optional*, defaults to `False`):
Whether or not to force and update of the internal state and display (by default, the bar will wait for
`value` to reach the value it predicted corresponds to a time of more than the `update_every` attribute
since the last update to avoid adding boilerplate).
comment (`str`, *optional*):
A comment to add on the left of the progress bar.
N)r" r@ r? time
start_time last_time start_value elapsed_timepredicted_remainingwarmup first_callswait_for
update_barminr# average_time_per_itemr6 closemaxr rD )rE r" rI r@ current_times &&&& r updateNotebookProgressBar.update s
"L ?? "/3yy{ :DOdn16 6D t;? ?D 8 #{{D DM OOE "
oo
%l
T - - 1Uc$//TXTaTaBacgcmcm>n5n !# A% 99;L , >D '''-1->->%JZJZBZ-[*-1*
"
+/(zzzJJL++7+/+E+EV[I[+\( OOE " #O )N**28R8RVW8W !
#C(9(9DDJ
%
%
- \N!DJJ4::,a D \ P ! \ P ! V P 4 RR7 V n R # V P P \ P ! V P 4 4 R # NT)
display_id)
r' r" r# r$ r% r& r. r7 rc rA dispHTMLrZ rE s &r rc NotebookProgressBar.display s *4::tzz4;;PTPZPZ\`\f\fg ;; " KK ! ;; ,,tyy '@TRDK KK tyy 8 9r c V P fB V P e2 V P P \ P ! R4 4 R# R# R# )zCloses the progress bar.Nr= )r7 rA rZ ri rj rk s &r rW NotebookProgressBar.close s7 ;; 4;;#: KK tyy} - $; r )rV r@ rO rR r. r% rM r? r6 rA r7 rP r$ rL rN r# rD r" rS r& )NTN, )FNr )__name__
__module__ __qualname____firstlineno____doc__rQ rD rF rZ rT rc rW __static_attributes____classdictcell__r: s @r r4 r4 F sB " H FL$ $,1\ 1\f& :. .r r4 c V a a ] tR t^t oRtRV 3R lltR tR tR R ltR t Rt
Vt V ;t # )
r8 a
An object tracking the updates of an ongoing training with progress bars and a nice table reporting metrics.
Args:
num_steps (`int`): The number of steps during training. column_names (`list[str]`, *optional*):
The list of column names for the metrics table (will be inferred from the first call to
[`~utils.notebook.NotebookTrainingTracker.write_line`] if not set).
c P < \ SV ` V4 Vf R MV.V n R V n R # r )superrF inner_table child_bar)rE num_steps column_names __class__s &&&r rF NotebookTrainingTracker.__init__ s(
##/#74l^ r c \ V P V P V P V P V P
4 V n V P e- V ;P \ V P 4 ,
un V P e. V ;P V P P ,
un V P f> \ P ! \ P ! V P 4 RR7 V n
R # V P P \ P ! V P 4 4 R # rg )r' r" r# r$ r% r& r. r{ r2 r| rA ri rc rj rZ rk s &r rc NotebookTrainingTracker.display s *4::tzz4;;PTPZPZ\`\f\fg ' NN 01A1A B BN >> % NNdnn 6 6 6N ;; ,,tyy '@TRDK KK tyy 8 9r c n V P f: \ VP 4 4 \ VP 4 4 .V n R# V P ^ , pV F pW29 g K VP V4 K W P ^ & \ V P 4 ^8 d V P R, pV P ^ , ^ , pV^ , W, 8w d= V P P V Uu. uF qaP
VR4 NK up4 R# TpV F$ pWg9 g K WBP V4 , Wv& K& V Uu. uF qgV, NK
upV P R& R# V P P V Uu. uF qaV, NK
up4 R# u upi u upi u upi )ze
Write the values in the inner table.
Args:
values (`dict[str, float]`): The values to display.
NzNo Log)r{ listkeysvaluesappendrb getindex)rE r columnskey last_values first_columnc
new_valuess && r
write_line"NotebookTrainingTracker.write_line si # $V[[] 3T&--/5J KD &&q)G %NN3' #* Q 4##$q( " . .r 2 # / / 21 5 q>V%99$$++g,VgZZ8-Dg,VW "(J $ .,7
a8H,IJM % DK+K7aqMM7+KD$$R( ''G(DGqG(DE -W ,L(Ds ?F(F-F2c > \ WWR7 V n V P # )a
Add a child progress bar displayed under the table of metrics. The child progress bar is returned (so it can be
easily updated).
Args:
total (`int`): The number of iterations for the child progress bar.
prefix (`str`, *optional*): A prefix to write on the left of the progress bar.
width (`int`, *optional*, defaults to 300): The width (in pixels) of the progress bar.
)r$ r7 r& )r4 r| )rE r# r$ r& s &&&&r add_child!NotebookTrainingTracker.add_child s -U$\~~ r c 4 RV n V P 4 R# )z
Closes the child progress bar.
N)r| rc rk s &r remove_child$NotebookTrainingTracker.remove_child! s r )r| r. r{ rA r )Nro )
rp rq rr rs rt rF rc r r r ru rv
__classcell__)r r: s @@r r8 r8 s)
: F> r r8 c ` a ] tR tRt o RtR tR tR tR
R ltR t R
R lt
R
R
lt R t R t
V tR# )NotebookProgressCallbacki) z~
A [`TrainerCallback`] that displays the progress of training or evaluation, optimized for Jupyter Notebooks or
Google colab.
c 0 R V n R V n RV n R # )NF)training_trackerprediction_bar_force_next_updaterk s &r rF !NotebookProgressCallback.__init__/ s $ ""' r c 2 VP \ P 8X d R MRV n ^ V n ^ V n V P .R., pVP \ P 8w d VP R4 \ VP V4 V n
R# )EpochStep
Training LossValidation LossN)
eval_strategyr EPOCHr
training_losslast_logNOr r8 max_stepsr )rE argsstatecontrolkwargsr~ s &&&&, r on_train_begin'NotebookProgressCallback.on_train_begin4 s} '+'9'9=M=S=S'SGY_
))*o->> !1!4!4 4 1 2 7 V r c L \ VP 4 VP 8X d \ VP 4 M
VP R
p\ V P R4 pVP VP
^, RV RVP 2V P R7 RV n R# )r` z0on_train_begin must be called before on_step_endEpoch r_ r@ rI FN)r epochr r rZ global_stepnum_train_epochsr )rE r r r r r tts &&&&, r on_step_end$NotebookProgressCallback.on_step_end= s $' $4 $CEKK EKKX[K\
d++-_
`
! UG1U%;%;$<= 00
#( r Nc \ V4 ' g R # V P fo V P e+ V P P \ V4 4 V n M\ \ V4 4 V n V P P
^4 R # V P P
V P P ^, 4 R # r )r r r r rb r4 rZ r" )rE r r r eval_dataloaderr s &&&&&,r on_prediction_step+NotebookProgressCallback.on_prediction_stepG s /** &$$0&*&;&;&E&Ec/FZ&[#&9#o:N&O# & &q ) & &t':':'@'@1'D Er c d V P e V P P 4 R V n R # r )r rW )rE r r r r s &&&&,r
on_predict#NotebookProgressCallback.on_predictS s( * % % ' "r c VP \ P 8X dM R V9 dD \ V P R4 pRVR , /pVP
VR& VP
V4 R# R# R# )lossz+on_train_begin must be called before on_logr r N)r r r r r r r )rE r r r logsr r r s &&&&&, r on_log NotebookProgressCallback.on_logX sb !1!4!4 44 $//1^_B%tF|4F " . .F6N MM& ! :H 4r c VP \ P 8X d R MRV n RRRR/p\ VP
4 F pRV9 g K VR, VR& M V P R 8X d \
VP 4 VR & MVP VR& Vf / pRpV F4 p V P R4 ' g K \ P ! R R
V 4 pK6 VP R R4 p
VP R R4 p
VP V R
2R4 p
VP V R2R4 p
VP V R2R4 p
VP V R2R4 p
VP 4 FX w rV P R4 p RP V R, U
u. uF qP! 4 NK up
4 pVR8X d RpWV& KZ V P" e6 V P" pVP% V4 VP' 4 RV n Mc\* P, ! \* P. ! \1 \3 VP5 4 4 \3 VP7 4 4 .4 4 4 RV n R# u up
i )r r r zNo logr r Neval_lossz\_loss$r=
total_flosr _runtime_samples_per_second_steps_per_second_model_preparation_time_r] r) LossT) r r r r reversed log_historyr r r endswithresubpopr- splitjoin
capitalizer r r r ri rc rj r2 r r r r )rE r r r metricsr r logmetric_key_prefixkr vsplitspartnamer s &&&&&, r on_evaluate$NotebookProgressCallback.on_evaluatea s. (,'9'9=M=S=S'SGY_!8->IE - -.C}*-f+' /
' !%++ .F7O " . .F6N ?G "Azz'""$&FF:r1$=!
KK d + KK & KK,-X6 = KK,--@A4 H KK,-->? F KK,--DEt LMMODAWWS\F886": F:4__ .: FGDv~( 4L
$ ,&&B MM& ! OO &*D #
LL#5tFKKM7JDQWQ^Q^Q`La6b#cd e "! Gs I.
c \ V P R 4 pVP VP R\ VP
4 RVP 2RR7 RV n R# )z1on_train_begin must be called before on_train_endr r_ Tr N)r r rZ r r r r )rE r r r r r s &&&&, r on_train_end%NotebookProgressCallback.on_train_end s]
d++-`
a
S - .a0F0F/GH
!% r )r r r r r r r )rp rq rr rs rt rF r r r r r r r ru rv rw s @r r r ) s:
(
W(
F#
".#`% %r r )ro )rB r rK typingr r IPython.displayrc ri trainer_callbackr
trainer_utilsr r r r r r' r2 r4 r8 r r! r r r sg
$ . 8
T]
D"R. R.jK 1 K\o% o%r