AVhdZddlZddlmZddlmZddlmZddlm Z ddl m Z ddlmZdd l mZdd l mZdd l mZdd lmZdd lmZddlmZddlmZm Z m!Z!ddl"m#Z#ejHdgdZ%d>de#e!ejLfde'de'de(de'de'f dZ)edejTe)Z+de#e!ejLfde'de'de(de'de'f dZ,ejHdgdZ-d>de#e!ejLfde#e!ejLfde'de'de'f dZ.ed ejTe.Z/de#e!ejLfde#e!ejLfde'de'de'f d!Z0ejHd"gdZ1d#d$dd$dgdddf de#e!ejLfde'de'de(d%e'd&e2d'e3d(e'd)e'd*e'de'de'fd+Z4ed,ejTe4Z5de#e!ejLfde'de'de(d%e'd&e2d'e3d(e'd)e'd*e'de'de'fd-Z6ejHd.gdZ7d>de#e!ejLfde'de'de(d%e'de'de'fd/Z8ed0ejTe8Z9de#e!ejLfde'de'de(d%e'de'de'fd1Z:ejHd2gdZ;d>de#e!ejLfde'de'de(d%e'de'de'fd3Z<ed4ejTe<Z=de#e!ejLfde'de'de(d%e'de'de'fd5Z>ejHd6gdZ?d>de#e!ejLfde'de'de(d%e'de'de'fd7Z@ed8ejTe@ZAde#e!ejLfde'de'de(d%e'de'de'fd9ZBejHd:gdZCd>de#e!ejLfde'de'de(d%e'de'de'fd;ZDed<ejTeDZEde#e!ejLfde'de'de(d%e'de'de'fd=ZFy)?zUPython wrappers around TensorFlow ops. This file is MACHINE GENERATED! Do not edit. N) pywrap_tfe)context)core)execute)dtypes)annotation_types)op_def_registry)ops)op_def_library)deprecated_endpoints)dispatch) tf_export)TypeVarListAny) AnnotatedAllCandidateSampler)sampled_candidatestrue_expected_countsampled_expected_count true_classesnum_true num_sampleduniqueseedseed2chtjxstj}|j}|jr: t j |d||d|d|d|d|d|} t j| } | Stj |d}tj |d}tj"|d}|d }tj |d}|d }tj |d}t%j&d||||||| \} } } } | dd} tj(r{d| j+dd| j+dd| j-dd| j+dd| j+df }| j.}tj0d||| t j| } | S#tj$r } tj| |Yd} ~ nd} ~ wtj$rYnwxYw t||||||||S#tj$rYwxYw) aGenerates labels for candidate sampling with a learned unigram distribution. See explanations of candidate sampling and the data formats at go/candidate-sampling. For each batch, this op picks a single set of sampled candidate labels. The advantages of sampling candidates per-batch are simplicity and the possibility of efficient dense matrix multiplication. The disadvantage is that the sampled candidates must be chosen independently of the context and of the true labels. Args: true_classes: A `Tensor` of type `int64`. A batch_size * num_true matrix, in which each row contains the IDs of the num_true target_classes in the corresponding original label. num_true: An `int` that is `>= 1`. Number of true labels per context. num_sampled: An `int` that is `>= 1`. Number of candidates to produce. unique: A `bool`. If unique is true, we sample with rejection, so that all sampled candidates in a batch are unique. This requires some approximation to estimate the post-rejection sampling probabilities. seed: An optional `int`. Defaults to `0`. If either seed or seed2 are set to be non-zero, the random number generator is seeded by the given seed. Otherwise, it is seeded by a random seed. seed2: An optional `int`. Defaults to `0`. An second seed to avoid seed collision. name: A name for the operation (optional). Returns: A tuple of `Tensor` objects (sampled_candidates, true_expected_count, sampled_expected_count). sampled_candidates: A `Tensor` of type `int64`. true_expected_count: A `Tensor` of type `float32`. sampled_expected_count: A `Tensor` of type `float32`. rrrrrrN)rrrrrnamectxr)rrrrrrr)_contextr_thread_local_datais_eagerrTFE_Py_FastPathExecute_AllCandidateSamplerOutput_make_core_NotOkStatusException_opsraise_from_not_ok_status_FallbackException$all_candidate_sampler_eager_fallback_SymbolicException_executemake_int make_bool_op_def_library_apply_op_helpermust_record_gradient _get_attr_int_get_attr_boolinputsrecord_gradient)rrrrrrr_ctxtld_resulte__op_outputs_attrs _inputs_flats `/home/dcms/DCMS/lib/python3.12/site-packages/tensorflow/python/ops/gen_candidate_sampling_ops.pyall_candidate_samplerrAs:L    0h..0$ #\\  11 #T<X{HffdG g+009g n  x 4(!!+}=+   fh /& \ D   4 ($ ] E   E7 +%'88L8+6v$(DB!QX QK' ""$#++J7 .  *FC4E4Ef4Ms((13F::L |VW> & , ,W 5' .E  & &- ##At,,  # #    1 {d%dFF  # #   s08F<<HG**HHHH10H1zraw_ops.AllCandidateSamplerc tj|d}tj|d}tj|d}|d}tj|d}|d}tj|d}tj|t j }|g}d|d|d|d|d|f } tjdd|| || } tjrtjd || | tj| } | S) NrrrrrrsAllCandidateSamplerr5attrsrrr) r-r.r/r(convert_to_tensor_dtypesint64rr2r6r$r%) rrrrrrrrr?r>r9s r@r+r+ts   x 4(!!+}=+   fh /& \ D   4 ($ ] E   E7 +%'' gmmD,, -h&$ (&   3Q|#)s ?' ""$ |VW> & , ,W 5' .ComputeAccidentalHits)indicesidsweightsrc tjxstj}|j}|jr7 t j |d|||d|d|d| }t j|}|Stj |d}|d}tj |d}|d}tj |d}t#j$d||||||\} } } } | dd}tj&rYd| j)dd| j)dd| j)df} | j*}tj,d|| |t j|}|S#tj$r } tj| |Yd} ~ nd} ~ wtj$rYnwxYw t|||||||S#tj$rYgwxYw) aComputes the ids of the positions in sampled_candidates that match true_labels. When doing log-odds NCE, the result of this op should be passed through a SparseToDense op, then added to the logits of the sampled candidates. This has the effect of 'removing' the sampled labels that match the true labels by making the classifier sure that they are sampled labels. Args: true_classes: A `Tensor` of type `int64`. The true_classes output of UnpackSparseLabels. sampled_candidates: A `Tensor` of type `int64`. The sampled_candidates output of CandidateSampler. num_true: An `int`. Number of true labels per context. seed: An optional `int`. Defaults to `0`. If either seed or seed2 are set to be non-zero, the random number generator is seeded by the given seed. Otherwise, it is seeded by a random seed. seed2: An optional `int`. Defaults to `0`. An second seed to avoid seed collision. name: A name for the operation (optional). Returns: A tuple of `Tensor` objects (indices, ids, weights). indices: A `Tensor` of type `int32`. ids: A `Tensor` of type `int64`. weights: A `Tensor` of type `float32`. rJrrrN)rrrrrr)rrrrrr)r rr!r"rr#_ComputeAccidentalHitsOutputr%r&r'r(r)r*&compute_accidental_hits_eager_fallbackr,r-r.r0r1r2r3r5r6)rrrrrrr7r8r9r:r;r<r=r>r?s r@compute_accidental_hitsrQs:    0h..0$ #\\  11 %t\;MHfdGUr9s r@rPrPs   x 4( \ D   4 ($ ] E   E7 +%'' gmmD,--.@'--P 23, &$ ?&   5q#)s ?' ""$ vw@ ( . .w 7' .rIFixedUnigramCandidateSampler range_max vocab_file distortionnum_reserved_ids num_shardsshardc^tjxstj}|j}|jrH t j |d| |d|d|d|d|d|d|d|d |d | d | d | d | }t j|}|Stj |d}tj |d}tj"|d}tj |d}|d}tj$|d}|d}tj&|d}|d}tj |d}|d}tj |d }| d} tj | d } | g} t)| t*t,fst/d| z| Dcgc]}tj&|d } }| d} tj | d } | d} tj | d } t1j2d|||||||||| | | | | \}}}}|dd}tj4rd|j7dd|j7dd|j9dd|j7dd|j;dd|j;dd|j7dd |j7d d |j7d d |j;d d |j7d d |j7d f}|j<}tj>d|||t j|}|S#tj$r }tj|| Yd}~nd}~wtj$rYnwxYw t|||||||||| | | | | |S#tj$rYwxYwcc}w)aAGenerates labels for candidate sampling with a learned unigram distribution. A unigram sampler could use a fixed unigram distribution read from a file or passed in as an in-memory array instead of building up the distribution from data on the fly. There is also an option to skew the distribution by applying a distortion power to the weights. The vocabulary file should be in CSV-like format, with the last field being the weight associated with the word. For each batch, this op picks a single set of sampled candidate labels. The advantages of sampling candidates per-batch are simplicity and the possibility of efficient dense matrix multiplication. The disadvantage is that the sampled candidates must be chosen independently of the context and of the true labels. Args: true_classes: A `Tensor` of type `int64`. A batch_size * num_true matrix, in which each row contains the IDs of the num_true target_classes in the corresponding original label. num_true: An `int` that is `>= 1`. Number of true labels per context. num_sampled: An `int` that is `>= 1`. Number of candidates to randomly sample. unique: A `bool`. If unique is true, we sample with rejection, so that all sampled candidates in a batch are unique. This requires some approximation to estimate the post-rejection sampling probabilities. range_max: An `int` that is `>= 1`. The sampler will sample integers from the interval [0, range_max). vocab_file: An optional `string`. Defaults to `""`. Each valid line in this file (which should have a CSV-like format) corresponds to a valid word ID. IDs are in sequential order, starting from num_reserved_ids. The last entry in each line is expected to be a value corresponding to the count or relative probability. Exactly one of vocab_file and unigrams needs to be passed to this op. distortion: An optional `float`. Defaults to `1`. The distortion is used to skew the unigram probability distribution. Each weight is first raised to the distortion's power before adding to the internal unigram distribution. As a result, distortion = 1.0 gives regular unigram sampling (as defined by the vocab file), and distortion = 0.0 gives a uniform distribution. num_reserved_ids: An optional `int`. Defaults to `0`. Optionally some reserved IDs can be added in the range [0, ..., num_reserved_ids) by the users. One use case is that a special unknown word token is used as ID 0. These IDs will have a sampling probability of 0. num_shards: An optional `int` that is `>= 1`. Defaults to `1`. A sampler can be used to sample from a subset of the original range in order to speed up the whole computation through parallelism. This parameter (together with 'shard') indicates the number of partitions that are being used in the overall computation. shard: An optional `int` that is `>= 0`. Defaults to `0`. A sampler can be used to sample from a subset of the original range in order to speed up the whole computation through parallelism. This parameter (together with 'num_shards') indicates the particular partition number of a sampler op, when partitioning is being used. unigrams: An optional list of `floats`. Defaults to `[]`. A list of unigram counts or probabilities, one per ID in sequential order. Exactly one of vocab_file and unigrams should be passed to this op. seed: An optional `int`. Defaults to `0`. If either seed or seed2 are set to be non-zero, the random number generator is seeded by the given seed. Otherwise, it is seeded by a random seed. seed2: An optional `int`. Defaults to `0`. An second seed to avoid seed collision. name: A name for the operation (optional). Returns: A tuple of `Tensor` objects (sampled_candidates, true_expected_count, sampled_expected_count). sampled_candidates: A `Tensor` of type `int64`. true_expected_count: A `Tensor` of type `float32`. sampled_expected_count: A `Tensor` of type `float32`. rSrrrrVrWrXrYrZr[unigramsrrN)rrrrVrWrXrYrZr[r]rrrrrTrUrVExpected list for 'unigrams' argument to 'fixed_unigram_candidate_sampler' Op, not %r.)rrrrrVrWrXrYrZr[r]rrr) r rr!r"rr##_FixedUnigramCandidateSamplerOutputr%r&r'r(r)r*.fixed_unigram_candidate_sampler_eager_fallbackr,r-r.r/make_str make_float isinstancelisttuple TypeErrorr0r1r2r3r4get_attrr5r6)rrrrrVrWrXrYrZr[r]rrrr7r8r9r:_fr;r<r=r>r?s r@fixed_unigram_candidate_samplerrisV    0h..0$ #\\  11 ,dL*-h <\:,lJ HfdGU Lg 499'Bg n  x 4(!!+}=+   fh /& ;7)J  \:*J"":|<*&&'79KLJ  \:* ] E   E7 +% H HtUm ,  8:B C DD=E Ebh!!"j1 E( E \ D   4 ($ ] E   E7 +%'88&\194?/53=3=9I3=U19.3$ @!QX QK' ""$#++J7 .  *K ,lll<(,ll<(*< 23\ -w(*ll:&0A0A&0Is((1 3F::L & fgG / 5 5g >' .U  & &- ##At,,  # #    ; {92BuxdDd ,,  # #   8Fs7AL+ N*+M2>MM21M26NN'&N'z$raw_ops.FixedUnigramCandidateSamplerctj|d}tj|d}tj|d}tj|d}|d}tj|d}|d}tj|d}|d }tj|d }|d}tj|d }| d } tj| d } | g} t | t tfstd | z| Dcgc]}tj|d} }| d } tj| d} | d } tj| d} tj|tj}|g}d|d|d|d|d|d|d |d |d | d| d| d| f}tjdd|||| }tjrtjd|||t j#|}|Scc}w)NrrrrVrTrWrUrXrrYrZr[r^r]rrsFixedUnigramCandidateSamplerrCrDrS)r-r.r/rarbrcrdrerfr(rFrGrHrr2r6r_r%)rrrrrVrWrXrYrZr[r]rrrrrhr?r>r9s r@r`r`s,   x 4(!!+}=+   fh /& ;7)J  \:*J"":|<*&&'79KLJ  \:* ] E   E7 +% H HtUm ,  8:B C DD=E Ebh!!"j1 E( E \ D   4 ($ ] E   E7 +%'' gmmD,, -h+y, L  "2L* 5*hgu F&   ' .)FsG9LearnedUnigramCandidateSamplerctjxstj}|j} | jr< t j |d||d|d|d|d|d|d|} t j| } | Stj |d}tj |d}tj"|d}tj |d}|d }tj |d}|d }tj |d}t%j&d|||||||| \} } } }|dd} tj(rd| j+dd| j+dd| j-dd| j+dd| j+dd| j+df }| j.}tj0d||| t j| } | S#tj$r } tj| |Yd} ~ nd} ~ wtj$rYnwxYw t||||||||| S#tj$rYwxYw) %Generates labels for candidate sampling with a learned unigram distribution. See explanations of candidate sampling and the data formats at go/candidate-sampling. For each batch, this op picks a single set of sampled candidate labels. The advantages of sampling candidates per-batch are simplicity and the possibility of efficient dense matrix multiplication. The disadvantage is that the sampled candidates must be chosen independently of the context and of the true labels. Args: true_classes: A `Tensor` of type `int64`. A batch_size * num_true matrix, in which each row contains the IDs of the num_true target_classes in the corresponding original label. num_true: An `int` that is `>= 1`. Number of true labels per context. num_sampled: An `int` that is `>= 1`. Number of candidates to randomly sample. unique: A `bool`. If unique is true, we sample with rejection, so that all sampled candidates in a batch are unique. This requires some approximation to estimate the post-rejection sampling probabilities. range_max: An `int` that is `>= 1`. The sampler will sample integers from the interval [0, range_max). seed: An optional `int`. Defaults to `0`. If either seed or seed2 are set to be non-zero, the random number generator is seeded by the given seed. Otherwise, it is seeded by a random seed. seed2: An optional `int`. Defaults to `0`. An second seed to avoid seed collision. name: A name for the operation (optional). Returns: A tuple of `Tensor` objects (sampled_candidates, true_expected_count, sampled_expected_count). sampled_candidates: A `Tensor` of type `int64`. true_expected_count: A `Tensor` of type `float32`. sampled_expected_count: A `Tensor` of type `float32`. rkrrrrVrrNrrrrVrrrrrrrrrrVrrr)r rr!r"rr#%_LearnedUnigramCandidateSamplerOutputr%r&r'r(r)r*0learned_unigram_candidate_sampler_eager_fallbackr,r-r.r/r0r1r2r3r4r5r6rrrrrVrrrr7r8r9r:r;r<r=r>r?s r@!learned_unigram_candidate_samplerrssgR    0h..0$ #\\  11 .lHm[(FYgu>g6;;GDg n  x 4(!!+}=+   fh /& ;7) \ D   4 ($ ] E   E7 +%'88(|3;6A179/35t M!QX QK' ""$#++J7 .  *K ,f'#2C2CG2L NF ::L (,I 1 7 7 @' .O  & &- ##At,,  # #    = {94u  # #   0:G&&H-9HH-,H-1IIIz&raw_ops.LearnedUnigramCandidateSamplerc @tj|d}tj|d}tj|d}tj|d}|d}tj|d}|d}tj|d}tj|t j }|g} d|d|d|d|d|d|f } tjdd | | || } tjrtjd | | | tj| } | S) NrrrrVrrrsLearnedUnigramCandidateSamplerrCrDrk) r-r.r/r(rFrGrHrr2r6rpr% rrrrrVrrrrr?r>r9s r@rqrq,s   x 4(!!+}=+   fh /& ;7) \ D   4 ($ ] E   E7 +%'' gmmD,, -h+y&$ @&   >$0C"& ('""$ (,I 1 7 7 @' .rILogUniformCandidateSamplerctjxstj}|j} | jr< t j |d||d|d|d|d|d|d|} t j| } | Stj |d}tj |d}tj"|d}tj |d}|d }tj |d}|d }tj |d}t%j&d|||||||| \} } } }|dd} tj(rd| j+dd| j+dd| j-dd| j+dd| j+dd| j+df }| j.}tj0d||| t j| } | S#tj$r } tj| |Yd} ~ nd} ~ wtj$rYnwxYw t||||||||| S#tj$rYwxYw) a!Generates labels for candidate sampling with a log-uniform distribution. See explanations of candidate sampling and the data formats at go/candidate-sampling. For each batch, this op picks a single set of sampled candidate labels. The advantages of sampling candidates per-batch are simplicity and the possibility of efficient dense matrix multiplication. The disadvantage is that the sampled candidates must be chosen independently of the context and of the true labels. Args: true_classes: A `Tensor` of type `int64`. A batch_size * num_true matrix, in which each row contains the IDs of the num_true target_classes in the corresponding original label. num_true: An `int` that is `>= 1`. Number of true labels per context. num_sampled: An `int` that is `>= 1`. Number of candidates to randomly sample. unique: A `bool`. If unique is true, we sample with rejection, so that all sampled candidates in a batch are unique. This requires some approximation to estimate the post-rejection sampling probabilities. range_max: An `int` that is `>= 1`. The sampler will sample integers from the interval [0, range_max). seed: An optional `int`. Defaults to `0`. If either seed or seed2 are set to be non-zero, the random number generator is seeded by the given seed. Otherwise, it is seeded by a random seed. seed2: An optional `int`. Defaults to `0`. An second seed to avoid seed collision. name: A name for the operation (optional). Returns: A tuple of `Tensor` objects (sampled_candidates, true_expected_count, sampled_expected_count). sampled_candidates: A `Tensor` of type `int64`. true_expected_count: A `Tensor` of type `float32`. sampled_expected_count: A `Tensor` of type `float32`. rwrrrrVrrNrnrro)r rr!r"rr#!_LogUniformCandidateSamplerOutputr%r&r'r(r)r*,log_uniform_candidate_sampler_eager_fallbackr,r-r.r/r0r1r2r3r4r5r6rrs r@log_uniform_candidate_samplerr{IsfR    0h..0$ #\\  11 *D, -h 64%1g277@g n  x 4(!!+}=+   fh /& ;7) \ D   4 ($ ] E   E7 +%'88$!QX QK' ""$#++J7 .  *K ,f'#2C2CG2L NF ::L $lFGE - 3 3G <' .O  & &- ##At,,  # #    9 {94u  # #   rtz"raw_ops.LogUniformCandidateSamplerc @tj|d}tj|d}tj|d}tj|d}|d}tj|d}|d}tj|d}tj|t j }|g} d|d|d|d|d|d|f } tjdd | | || } tjrtjd | | | tj| } | S) NrrrrVrrrsLogUniformCandidateSamplerrCrDrw) r-r.r/r(rFrGrHrr2r6ryr%rvs r@rzrzs   x 4(!!+}=+   fh /& ;7) \ D   4 ($ ] E   E7 +%'' gmmD,, -h+y&$ @&   :A$0C"& ('""$ $lFGE - 3 3G <' .rI#ThreadUnsafeUnigramCandidateSamplerctjxstj}|j} | jr< t j |d||d|d|d|d|d|d|} t j| } | Stj |d}tj |d}tj"|d}tj |d}|d }tj |d}|d }tj |d}t%j&d|||||||| \} } } }|dd} tj(rd| j+dd| j+dd| j-dd| j+dd| j+dd| j+df }| j.}tj0d||| t j| } | S#tj$r } tj| |Yd} ~ nd} ~ wtj$rYnwxYw t||||||||| S#tj$rYwxYw) rmr}rrrrVrrNrnrro)r rr!r"rr#*_ThreadUnsafeUnigramCandidateSamplerOutputr%r&r'r(r)r*6thread_unsafe_unigram_candidate_sampler_eager_fallbackr,r-r.r/r0r1r2r3r4r5r6rrs r@'thread_unsafe_unigram_candidate_samplerrshR    0h..0$ #\\  11 3T<Hm[(FYgu>g;@@Ig n  x 4(!!+}=+   fh /& ;7) \ D   4 ($ ] E   E7 +%'88-L8@;F6<9B5: G!QX QK' ""$#++J7 .  *K ,f'#2C2CG2L NF ::L -|VWN 6 < = 1`. Number of true labels per context. num_sampled: An `int` that is `>= 1`. Number of candidates to randomly sample. unique: A `bool`. If unique is true, we sample with rejection, so that all sampled candidates in a batch are unique. This requires some approximation to estimate the post-rejection sampling probabilities. range_max: An `int` that is `>= 1`. The sampler will sample integers from the interval [0, range_max). seed: An optional `int`. Defaults to `0`. If either seed or seed2 are set to be non-zero, the random number generator is seeded by the given seed. Otherwise, it is seeded by a random seed. seed2: An optional `int`. Defaults to `0`. An second seed to avoid seed collision. name: A name for the operation (optional). Returns: A tuple of `Tensor` objects (sampled_candidates, true_expected_count, sampled_expected_count). sampled_candidates: A `Tensor` of type `int64`. true_expected_count: A `Tensor` of type `float32`. sampled_expected_count: A `Tensor` of type `float32`. rrrrrVrrNrnrro)r rr!r"rr#_UniformCandidateSamplerOutputr%r&r'r(r)r*(uniform_candidate_sampler_eager_fallbackr,r-r.r/r0r1r2r3r4r5r6rrs r@uniform_candidate_samplerrBsfR    0h..0$ #\\  11 '|Z-h 64%1g/44W=g n  x 4(!!+}=+   fh /& ;7) \ D   4 ($ ] E   E7 +%'88! ,4+*0I(,E F!QX QK' ""$#++J7 .  *K ,f'#2C2CG2L NF ::L !<B * 0 0 9' .M  & &- ##At,,  # #    5 {94u  # #   rtzraw_ops.UniformCandidateSamplerc @tj|d}tj|d}tj|d}tj|d}|d}tj|d}|d}tj|d}tj|t j }|g} d|d|d|d|d|d|f } tjdd | | || } tjrtjd | | | tj| } | S) NrrrrVrrrsUniformCandidateSamplerrCrDr) r-r.r/r(rFrGrHrr2r6rr%rvs r@rrs   x 4(!!+}=+   fh /& ;7) \ D   4 ($ ] E   E7 +%'' gmmD,, -h+y&$ @&   7$0C"& ('""$ !<B * 0 0 9' .rI)rrN)G__doc__ collectionstensorflow.pythonrtensorflow.python.eagerrr rr&rr-tensorflow.python.frameworkrrGtensorflow.security.fuzzing.pyr_atypesr _op_def_registryr r(r r0"tensorflow.python.util.deprecationr tensorflow.python.utilr _dispatch tensorflow.python.util.tf_exportrtypingrrrtyping_extensionsr namedtupler$Int64intboolrA to_raw_oprr+rOrQrJrPr_strfloatrirSr`rprsrkrqryr{rwrzrrr}rrrrrrIr@rs 67179FK3IC86%%'3[33KM R #w}}2D(ERQTRcfRptR{~RILRh?i =>~t~~Nc?deygmmAS7T`cru@DLOX[, 6{55! # E)C4F*GE]fgjlslylygy]zEGJEQTE^aENC "AB>4>>RiCjk3 CU9Vluvy|C|I|IwImJVYadmp('=k&<&<"K'M# `buvMN_`lmxzEFRSZ^a)C(>$K)M% ZIc7==>P4QZ]`ZorZ}AZNQZX[ZehZx"U+S!TUcUYUcUceFVG"H9SRYR_R_M_C`loBLP]`hktw0%;K$:$: K%M! Z #w}}:L0MZY\ZknZx|ZJMZTWZadZxMY'KL^T^^\yMz{ygmmI[?\hkz}HLY\dgps0.D[-C-C)K.M* [)CDV:W[cf[ux[CG[TW[^a[kn[z'_i0]&^_m_c_m_moV`W'X#SVX_XeXeSeIfruEHRVcfnqz}0"8!7!7K"M YIc7==6H,IYUXYgjYtxYFIYPSY]`YvG)$EF~t~~VoGpq9S'--EW;XdgvyDHUX`clorI