Received: from malur.postgresql.org ([217.196.149.56]) by arkaria.postgresql.org with esmtps (TLS1.3:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.92) (envelope-from ) id 1kVFJi-0007FT-3d for pgsql-performance@arkaria.postgresql.org; Wed, 21 Oct 2020 14:42:18 +0000 Received: from localhost ([127.0.0.1] helo=malur.postgresql.org) by malur.postgresql.org with esmtp (Exim 4.92) (envelope-from ) id 1kVFJf-00075X-0B for pgsql-performance@arkaria.postgresql.org; Wed, 21 Oct 2020 14:42:15 +0000 Received: from magus.postgresql.org ([2a02:c0:301:0:ffff::29]) by malur.postgresql.org with esmtps (TLS1.3:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.92) (envelope-from ) id 1kVFJe-00073G-Gr for pgsql-performance@lists.postgresql.org; Wed, 21 Oct 2020 14:42:14 +0000 Received: from mail-lj1-x22a.google.com ([2a00:1450:4864:20::22a]) by magus.postgresql.org with esmtps (TLS1.3:ECDHE_RSA_AES_128_GCM_SHA256:128) (Exim 4.92) (envelope-from ) id 1kVFJa-0007Ge-Th for pgsql-performance@postgresql.org; Wed, 21 Oct 2020 14:42:13 +0000 Received: by mail-lj1-x22a.google.com with SMTP id a4so2859148lji.12 for ; Wed, 21 Oct 2020 07:42:10 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=duneanalytics-com.20150623.gappssmtp.com; s=20150623; h=subject:to:cc:references:from:message-id:date:user-agent :mime-version:in-reply-to:content-language; bh=uqIDYB5VWPeHuqrN1kVNfD8JkzXuFvtBzvlHvmLx11A=; b=G6VKlpMK/5SZet2VjMx0+k/D6kGiC6rnve/eifY8tvOe8deLM8X11frfSE5j0wieGk SmiLo1ihaDr0ijMWsOiruW4oWYzu+l2jkg8AwXM5VROvgL3A2hOJpHoNtsiMESikljlA xew9rLjDhdFjNI6SgB5Q6WikgaapC2R/6IXYhcQ1CAxYvlZjBISD3JoA0+QFPKgc068i meuKSd9OFumYljZCN7XxoGuN2NlBsOTPkYnMjEjj4MengJwXgaQz0pfBC59Y5ZXGdJsN dFM7Q+RYvbIN1QjlwzoUyZGpW6tISX3mrdRwAJgvazvIDK1tTiljUcKf9UxRhXXSkmmu VhAg== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:subject:to:cc:references:from:message-id:date :user-agent:mime-version:in-reply-to:content-language; bh=uqIDYB5VWPeHuqrN1kVNfD8JkzXuFvtBzvlHvmLx11A=; b=ic/yunZMsKN7EeDMKvPzq91UpbRnXWtGfx8sdypQkUid3YaHk00PVkOCrMJ3NWNEAn aaFtL7dFfMjW9QkQaVo7oNsypn+vfs01wbEH9hGyFriWWzMJ4AsbZJ/Sob8IWrgyn93j luz2NyjDbYloGNEuPi3CQIQTJewB0DrKbuLwuOUHDLuC+6CTsFFNdYkc1LrWALIukUkf sNzxb/zcxvt/m8+n5Gvkdu8wfBcIAHpfltrP+oafMynyVHWfKnCRKIKtCxWaMIErsYXH F5RLP3uekLBhtJZ6Z/g+FbcnMH3hX1hioEZ4GD1gZtr+O6l6KS/aXfRVEGXsqV3SEfXa 52zQ== X-Gm-Message-State: AOAM531qYASPog2jZR5QPFuz4+f/JZe8cWt85T9fXr5hh0CrKifStkQt pj70KHgdpQ7lQFqLrZmukU58z50epsY6Hted X-Google-Smtp-Source: ABdhPJy8yiS+8LNMkjAgWJakGAZlD3Cxy/19oFo4qhfmpPle0eS19ltPu+dK4+O64E67MTcso5j2DA== X-Received: by 2002:a2e:8515:: with SMTP id j21mr1616612lji.419.1603291329329; Wed, 21 Oct 2020 07:42:09 -0700 (PDT) Received: from [192.168.1.181] (host-37-191-238-241.lynet.no. [37.191.238.241]) by smtp.gmail.com with ESMTPSA id o21sm423018lff.265.2020.10.21.07.42.07 (version=TLS1_3 cipher=TLS_AES_128_GCM_SHA256 bits=128/128); Wed, 21 Oct 2020 07:42:08 -0700 (PDT) Subject: Re: Query Performance / Planner estimate off To: Sebastian Dressler Cc: "pgsql-performance@postgresql.org" References: From: Mats Olsen Message-ID: <61c9fd35-bdb8-3fe6-bda0-81c7e6a9eaf8@duneanalytics.com> Date: Wed, 21 Oct 2020 16:42:02 +0200 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101 Thunderbird/78.3.2 MIME-Version: 1.0 In-Reply-To: Content-Type: multipart/alternative; boundary="------------1578AA2B782F8AC1EEB779EB" Content-Language: en-US List-Id: List-Help: List-Subscribe: List-Post: List-Owner: List-Archive: Precedence: bulk This is a multi-part message in MIME format. --------------1578AA2B782F8AC1EEB779EB Content-Type: text/plain; charset=utf-8; format=flowed Content-Transfer-Encoding: 8bit On 10/21/20 2:38 PM, Sebastian Dressler wrote: > Hi Mats, > >> On 20. Oct 2020, at 11:37, Mats Julian Olsen > > wrote: >> >> [...] >> >> 1) Vanilla plan (16 min) : https://explain.depesz.com/s/NvDR >> >> 2) enable_nestloop=off (4 min): https://explain.depesz.com/s/buKK >> >> 3) enable_nestloop=off; enable_seqscan=off (2 min): >> https://explain.depesz.com/s/0WXx >> >> How can I get Postgres not to loop over 12M rows? > > I looked at the plans and your config and there are some thoughts I'm > having: > > - The row estimate is off, as you possibly noticed. This can be > possibly solved by raising `default_statistics_target` to e.g. 2500 > (we typically use that) and run ANALYZE I've `set default_statistics_target=2500` and ran analyze on both tables involved, unfortunately the plan is the same. The columns we use for joining here are hashes and we expect very few duplicates in the tables. Hence I think extended statistics (storing most common values and histogram bounds) aren't useful for this kind of data. Would you say the same thing? > > - I however think that the misestimate might be caused by the > evt_tx_hash being of type bytea. I believe that PG cannot estimate > this very well for JOINs and will rather pick row numbers too low. > Hence the nested loop is picked and there might be no way around this. > I have experienced similar things when applying JOINs on VARCHAR with > e.g. more than 3 fields for comparison. This is very interesting, and I have never heard of issues with using `bytea` for joins. Our entire database is filled with them, as we deal with hashes of different lengths. In fact I would estimate that 60% of columns are bytea's. My intuition would say that it's better to store the hashes as byte arrays, rather than `text` fields as you can compare the raw bytes directly without encoding first?  Do you have any references for this? Alternatively, since I know the length of the hashes in advance, I could've used `varchar(n)`, but I don't think there's any gains to be had in postgres by doing that? Something like `bytea(n)` would also have been interesting, had postgres been able to exploit that information. > - Other things to look into: > >     - work_mem seems too low to me with 56MB, consider raising this to > the GB range to avoid disk-based operations >     - min_parallel_table_scan_size - try 0 >     - parallel_setup_cost (default 1000, maybe try 500) >     - parallel_tuple_cost (default 1.0, maybe try 0.1) >     - random_page_cost (as mentioned consider raising this maybe much > higher, factor 10 or sth like this) or (typically) seq_page_cost can > be possibly much lower (0.1, 0.01) depending on your storage I've tried various settings of these parameters now, and unfortunately the only parameter that alters the query plan is the last one (random_page_cost), which also has the side effect of (almost) forcing sequential scans for most queries as far as I understand? Our storage is Google Cloud pd-ssd. Thank you so much for you response, I'm looking forward to keep the discussion going. > > I hope this helps to get to a parallel plan without setting > `nested_loop = off`. EXPLAIN should be enough already to see the > difference. > > Best, > Sebastian > > -- > > Sebastian Dressler, Solution Architect > +49 30 994 0496 72 | sebastian@swarm64.com > > Swarm64 AS > Parkveien 41 B | 0258 Oslo | Norway > Registered at Brønnøysundregistrene in Norway under Org.-Number 911 > 662 787 > CEO/Geschäftsführer (Daglig Leder): Thomas > Richter; Chairman/Vorsitzender (Styrets Leder): Dr. Sverre Munck > > Swarm64 AS Zweigstelle Hive > Ullsteinstr. 120 | 12109 Berlin | Germany > Registered at Amtsgericht Charlottenburg - HRB 154382 B > Best, Mats --------------1578AA2B782F8AC1EEB779EB Content-Type: multipart/related; boundary="------------5600935A196F9FFC789BF743" --------------5600935A196F9FFC789BF743 Content-Type: text/html; charset=utf-8 Content-Transfer-Encoding: 8bit


On 10/21/20 2:38 PM, Sebastian Dressler wrote:
Hi Mats,

On 20. Oct 2020, at 11:37, Mats Julian Olsen <mats@duneanalytics.com> wrote:

[...]

1) Vanilla plan (16 min) : https://explain.depesz.com/s/NvDR
2) enable_nestloop=off (4 min): https://explain.depesz.com/s/buKK
3) enable_nestloop=off; enable_seqscan=off (2 min): https://explain.depesz.com/s/0WXx 

How can I get Postgres not to loop over 12M rows?

I looked at the plans and your config and there are some thoughts I'm having:

- The row estimate is off, as you possibly noticed. This can be possibly solved by raising `default_statistics_target` to e.g. 2500 (we typically use that) and run ANALYZE
I've `set default_statistics_target=2500` and ran analyze on both tables involved, unfortunately the plan is the same. The columns we use for joining here are hashes and we expect very few duplicates in the tables. Hence I think extended statistics (storing most common values and histogram bounds) aren't useful for this kind of data. Would you say the same thing?

- I however think that the misestimate might be caused by the evt_tx_hash being of type bytea. I believe that PG cannot estimate this very well for JOINs and will rather pick row numbers too low. Hence the nested loop is picked and there might be no way around this. I have experienced similar things when applying JOINs on VARCHAR with e.g. more than 3 fields for comparison.

This is very interesting, and I have never heard of issues with using `bytea` for joins. Our entire database is filled with them, as we deal with hashes of different lengths. In fact I would estimate that 60% of columns are bytea's. My intuition would say that it's better to store the hashes as byte arrays, rather than `text` fields as you can compare the raw bytes directly without encoding first?  Do you have any references for this?

Alternatively, since I know the length of the hashes in advance, I could've used `varchar(n)`, but I don't think there's any gains to be had in postgres by doing that? Something like `bytea(n)` would also have been interesting, had postgres been able to exploit that information.


- Other things to look into:

    - work_mem seems too low to me with 56MB, consider raising this to the GB range to avoid disk-based operations
    - min_parallel_table_scan_size - try 0
    - parallel_setup_cost (default 1000, maybe try 500)
    - parallel_tuple_cost (default 1.0, maybe try 0.1)
    - random_page_cost (as mentioned consider raising this maybe much higher, factor 10 or sth like this) or (typically) seq_page_cost can be possibly much lower (0.1, 0.01) depending on your storage

I've tried various settings of these parameters now, and unfortunately the only parameter that alters the query plan is the last one (random_page_cost), which also has the side effect of (almost) forcing sequential scans for most queries as far as I understand? Our storage is Google Cloud pd-ssd.

Thank you so much for you response, I'm looking forward to keep the discussion going.


I hope this helps to get to a parallel plan without setting `nested_loop = off`. EXPLAIN should be enough already to see the difference.

Best,
Sebastian

--

Sebastian Dressler, Solution Architect 
+49 30 994 0496 72 | sebastian@swarm64.com 

Swarm64 AS
Parkveien 41 B | 0258 Oslo | Norway
Registered at Brønnøysundregistrene in Norway under Org.-Number 911 662 787
CEO/Geschäftsführer (Daglig Leder): Thomas Richter; Chairman/Vorsitzender (Styrets Leder): Dr. Sverre Munck 

Swarm64 AS Zweigstelle Hive
Ullsteinstr. 120 | 12109 Berlin | Germany
Registered at Amtsgericht Charlottenburg - HRB 154382 B 
 


Best,

Mats

--------------5600935A196F9FFC789BF743 Content-Type: image/jpeg; name="image001.jpg" Content-Transfer-Encoding: base64 Content-ID: Content-Disposition: inline; filename="image001.jpg" /9j/4QYpRXhpZgAATU0AKgAAAAgADAEAAAMAAAABAgkAAAEBAAMAAAABAI0AAAECAAMAAAAD AAAAngEGAAMAAAABAAIAAAESAAMAAAABAAEAAAEVAAMAAAABAAMAAAEaAAUAAAABAAAApAEb AAUAAAABAAAArAEoAAMAAAABAAIAAAExAAIAAAAkAAAAtAEyAAIAAAAUAAAA2IdpAAQAAAAB AAAA7AAAASQACAAIAAgALcbAAAAnEAAtxsAAACcQQWRvYmUgUGhvdG9zaG9wIENDIDIwMTkg KE1hY2ludG9zaCkAMjAxODoxMToxMyAxMTozMDoxOAAABJAAAAcAAAAEMDIyMaABAAMAAAAB AAEAAKACAAQAAAABAAAA/aADAAQAAAABAAAASgAAAAAAAAAGAQMAAwAAAAEABgAAARoABQAA AAEAAAFyARsABQAAAAEAAAF6ASgAAwAAAAEAAgAAAgEABAAAAAEAAAGCAgIABAAAAAEAAASf AAAAAAAAAEgAAAABAAAASAAAAAH/2P/tAAxBZG9iZV9DTQAB/+4ADkFkb2JlAGSAAAAAAf/b AIQADAgICAkIDAkJDBELCgsRFQ8MDA8VGBMTFRMTGBEMDAwMDAwRDAwMDAwMDAwMDAwMDAwM DAwMDAwMDAwMDAwMDAENCwsNDg0QDg4QFA4ODhQUDg4ODhQRDAwMDAwREQwMDAwMDBEMDAwM DAwMDAwMDAwMDAwMDAwMDAwMDAwMDAwM/8AAEQgAEgA9AwEiAAIRAQMRAf/dAAQABP/EAT8A AAEFAQEBAQEBAAAAAAAAAAMAAQIEBQYHCAkKCwEAAQUBAQEBAQEAAAAAAAAAAQACAwQFBgcI CQoLEAABBAEDAgQCBQcGCAUDDDMBAAIRAwQhEjEFQVFhEyJxgTIGFJGhsUIjJBVSwWIzNHKC 0UMHJZJT8OHxY3M1FqKygyZEk1RkRcKjdDYX0lXiZfKzhMPTdePzRieUpIW0lcTU5PSltcXV 5fVWZnaGlqa2xtbm9jdHV2d3h5ent8fX5/cRAAICAQIEBAMEBQYHBwYFNQEAAhEDITESBEFR YXEiEwUygZEUobFCI8FS0fAzJGLhcoKSQ1MVY3M08SUGFqKygwcmNcLSRJNUoxdkRVU2dGXi 8rOEw9N14/NGlKSFtJXE1OT0pbXF1eX1VmZ2hpamtsbW5vYnN0dXZ3eHl6e3x//aAAwDAQAC EQMRAD8A7nL62XZVmNWTXVU41l4ncXN+m5v8mt/t9nqKOP1S+hxNri9jHFtrHEOIgua5zH/2 f3v+t1rN6tjOovyKHn032PsuxrDO1wfue9m5vu9v/B+/+ux6P6L8sh2OJdkA7e4BJd6oft9v 6J/0/wDC/wAtYvNTzDIJxMhMT4QBda/LHg/dbfK8EuKE6oi9f+c6mT9Z+iY1uRXdkOH2Td9o sbVa+thY31rK35FdT8f1mVe99Pqeqiv690ivIfjPyWi+u+vEfXDiRdcz16atG/n1e/f/ADf/ AG2sy76oi7qd2Sb2VYeQLG5GNSy1jrhbS7DsbmO+1uw7vpNu3/YK7fUqp/SIOP8AUb0n4lx6 jaciiq5uTe1ga+7IsbeyjPkuf6d+H9uy/R/nv5yv/QraajpD619A+z2ZD8o1V0msWerXbU4e tIx3Gq6uu307tv6K3Z6b1F31w+rbbHVHNaLGl7SzZZu31nZZjtb6e5+U3/uIz9Z/wnpbFmYn 1HyKXtsvz2Wva7CJc2mwF32K85u6x2RmZT3XZW7bbZu2f4T0loY/1adRk4t/2ncMXPzM/Zsj d9sGS30Z9T2+h9s/nP8ACbP5utJTpY3UsLLqx7sW0X1ZbfUoezUFgE7/AOT+7/XU/tuN6hr3 iRYKZ7by3fsn5f56yOl/Vh3T8PFxTktt+zNa11oq2vlllmS11BNlnob/AFtl3896jFZHQ4Y6 r7VYanOaRozcA2t9Eb9u3e7fv9T0/wDwX9KoRLN7lGI9viPqv9Dh9P8Az2UxxcFifroemv0r 1/5r/9D0frH9Cd/ReR/Tf5n+0g9C/m7v6F9P/tDx/wBf/wCEXzSko/8AKfRT9VJL5VSUin6q SXyqkkp+qkl8qpJKf//Z/+0OMFBob3Rvc2hvcCAzLjAAOEJJTQQEAAAAAAAPHAFaAAMbJUcc AgAAAgAAADhCSU0EJQAAAAAAEM3P+n2ox74JBXB2rq8Fw044QklNBDoAAAAAAOUAAAAQAAAA AQAAAAAAC3ByaW50T3V0cHV0AAAABQAAAABQc3RTYm9vbAEAAAAASW50ZWVudW0AAAAASW50 ZQAAAABDbHJtAAAAD3ByaW50U2l4dGVlbkJpdGJvb2wAAAAAC3ByaW50ZXJOYW1lVEVYVAAA AAEAAAAAAA9wcmludFByb29mU2V0dXBPYmpjAAAADABQAHIAbwBvAGYAIABTAGUAdAB1AHAA AAAAAApwcm9vZlNldHVwAAAAAQAAAABCbHRuZW51bQAAAAxidWlsdGluUHJvb2YAAAAJcHJv b2ZDTVlLADhCSU0EOwAAAAACLQAAABAAAAABAAAAAAAScHJpbnRPdXRwdXRPcHRpb25zAAAA FwAAAABDcHRuYm9vbAAAAAAAQ2xicmJvb2wAAAAAAFJnc01ib29sAAAAAABDcm5DYm9vbAAA AAAAQ250Q2Jvb2wAAAAAAExibHNib29sAAAAAABOZ3R2Ym9vbAAAAAAARW1sRGJvb2wAAAAA AEludHJib29sAAAAAABCY2tnT2JqYwAAAAEAAAAAAABSR0JDAAAAAwAAAABSZCAgZG91YkBv 4AAAAAAAAAAAAEdybiBkb3ViQG/gAAAAAAAAAAAAQmwgIGRvdWJAb+AAAAAAAAAAAABCcmRU VW50RiNSbHQAAAAAAAAAAAAAAABCbGQgVW50RiNSbHQAAAAAAAAAAAAAAABSc2x0VW50RiNQ eGxAcsAAAAAAAAAAAAp2ZWN0b3JEYXRhYm9vbAEAAAAAUGdQc2VudW0AAAAAUGdQcwAAAABQ Z1BDAAAAAExlZnRVbnRGI1JsdAAAAAAAAAAAAAAAAFRvcCBVbnRGI1JsdAAAAAAAAAAAAAAA AFNjbCBVbnRGI1ByY0BZAAAAAAAAAAAAEGNyb3BXaGVuUHJpbnRpbmdib29sAAAAAA5jcm9w UmVjdEJvdHRvbWxvbmcAAAAAAAAADGNyb3BSZWN0TGVmdGxvbmcAAAAAAAAADWNyb3BSZWN0 UmlnaHRsb25nAAAAAAAAAAtjcm9wUmVjdFRvcGxvbmcAAAAAADhCSU0D7QAAAAAAEAEsAAAA AQACASwAAAABAAI4QklNBCYAAAAAAA4AAAAAAAAAAAAAP4AAADhCSU0EDQAAAAAABAAAAFo4 QklNBBkAAAAAAAQAAAAeOEJJTQPzAAAAAAAJAAAAAAAAAAABADhCSU0nEAAAAAAACgABAAAA AAAAAAI4QklNA/UAAAAAAEgAL2ZmAAEAbGZmAAYAAAAAAAEAL2ZmAAEAoZmaAAYAAAAAAAEA MgAAAAEAWgAAAAYAAAAAAAEANQAAAAEALQAAAAYAAAAAAAE4QklNA/gAAAAAAHAAAP////// //////////////////////8D6AAAAAD/////////////////////////////A+gAAAAA//// /////////////////////////wPoAAAAAP////////////////////////////8D6AAAOEJJ TQQAAAAAAAACAAA4QklNBAIAAAAAAAIAADhCSU0EMAAAAAAAAQEAOEJJTQQtAAAAAAAGAAEA AADWOEJJTQQIAAAAAABMAAAAAQAAAkAAAAJAAAAADAAAC8QB///9lAEAAANxAQAAAT4AAAEy 5QD///IvAf//7tEB///0eAEAAAq8Af//+AoBAAAJOwEAAAVlAThCSU0EHgAAAAAABAAAAAA4 QklNBBoAAAAAA0sAAAAGAAAAAAAAAAAAAABKAAAA/QAAAAsARgBvAG8AdABlAHIAIABMAG8A ZwBvAAAAAQAAAAAAAAAAAAAAAAAAAAAAAAABAAAAAAAAAAAAAAD9AAAASgAAAAAAAAAAAAAA AAAAAAABAAAAAAAAAAAAAAAAAAAAAAAAABAAAAABAAAAAAAAbnVsbAAAAAIAAAAGYm91bmRz T2JqYwAAAAEAAAAAAABSY3QxAAAABAAAAABUb3AgbG9uZwAAAAAAAAAATGVmdGxvbmcAAAAA AAAAAEJ0b21sb25nAAAASgAAAABSZ2h0bG9uZwAAAP0AAAAGc2xpY2VzVmxMcwAAAAFPYmpj AAAAAQAAAAAABXNsaWNlAAAAEgAAAAdzbGljZUlEbG9uZwAAAAAAAAAHZ3JvdXBJRGxvbmcA AAAAAAAABm9yaWdpbmVudW0AAAAMRVNsaWNlT3JpZ2luAAAADWF1dG9HZW5lcmF0ZWQAAAAA VHlwZWVudW0AAAAKRVNsaWNlVHlwZQAAAABJbWcgAAAABmJvdW5kc09iamMAAAABAAAAAAAA UmN0MQAAAAQAAAAAVG9wIGxvbmcAAAAAAAAAAExlZnRsb25nAAAAAAAAAABCdG9tbG9uZwAA AEoAAAAAUmdodGxvbmcAAAD9AAAAA3VybFRFWFQAAAABAAAAAAAAbnVsbFRFWFQAAAABAAAA AAAATXNnZVRFWFQAAAABAAAAAAAGYWx0VGFnVEVYVAAAAAEAAAAAAA5jZWxsVGV4dElzSFRN TGJvb2wBAAAACGNlbGxUZXh0VEVYVAAAAAEAAAAAAAlob3J6QWxpZ25lbnVtAAAAD0VTbGlj ZUhvcnpBbGlnbgAAAAdkZWZhdWx0AAAACXZlcnRBbGlnbmVudW0AAAAPRVNsaWNlVmVydEFs aWduAAAAB2RlZmF1bHQAAAALYmdDb2xvclR5cGVlbnVtAAAAEUVTbGljZUJHQ29sb3JUeXBl AAAAAE5vbmUAAAAJdG9wT3V0c2V0bG9uZwAAAAAAAAAKbGVmdE91dHNldGxvbmcAAAAAAAAA DGJvdHRvbU91dHNldGxvbmcAAAAAAAAAC3JpZ2h0T3V0c2V0bG9uZwAAAAAAOEJJTQQoAAAA AAAMAAAAAj/wAAAAAAAAOEJJTQQUAAAAAAAEAAAA1jhCSU0EDAAAAAAEuwAAAAEAAAA9AAAA EgAAALgAAAzwAAAEnwAYAAH/2P/tAAxBZG9iZV9DTQAB/+4ADkFkb2JlAGSAAAAAAf/bAIQA DAgICAkIDAkJDBELCgsRFQ8MDA8VGBMTFRMTGBEMDAwMDAwRDAwMDAwMDAwMDAwMDAwMDAwM DAwMDAwMDAwMDAENCwsNDg0QDg4QFA4ODhQUDg4ODhQRDAwMDAwREQwMDAwMDBEMDAwMDAwM DAwMDAwMDAwMDAwMDAwMDAwMDAwM/8AAEQgAEgA9AwEiAAIRAQMRAf/dAAQABP/EAT8AAAEF AQEBAQEBAAAAAAAAAAMAAQIEBQYHCAkKCwEAAQUBAQEBAQEAAAAAAAAAAQACAwQFBgcICQoL EAABBAEDAgQCBQcGCAUDDDMBAAIRAwQhEjEFQVFhEyJxgTIGFJGhsUIjJBVSwWIzNHKC0UMH JZJT8OHxY3M1FqKygyZEk1RkRcKjdDYX0lXiZfKzhMPTdePzRieUpIW0lcTU5PSltcXV5fVW ZnaGlqa2xtbm9jdHV2d3h5ent8fX5/cRAAICAQIEBAMEBQYHBwYFNQEAAhEDITESBEFRYXEi EwUygZEUobFCI8FS0fAzJGLhcoKSQ1MVY3M08SUGFqKygwcmNcLSRJNUoxdkRVU2dGXi8rOE w9N14/NGlKSFtJXE1OT0pbXF1eX1VmZ2hpamtsbW5vYnN0dXZ3eHl6e3x//aAAwDAQACEQMR AD8A7nL62XZVmNWTXVU41l4ncXN+m5v8mt/t9nqKOP1S+hxNri9jHFtrHEOIgua5zH/2f3v+ t1rN6tjOovyKHn032PsuxrDO1wfue9m5vu9v/B+/+ux6P6L8sh2OJdkA7e4BJd6oft9v6J/0 /wDC/wAtYvNTzDIJxMhMT4QBda/LHg/dbfK8EuKE6oi9f+c6mT9Z+iY1uRXdkOH2Td9osbVa +thY31rK35FdT8f1mVe99Pqeqiv690ivIfjPyWi+u+vEfXDiRdcz16atG/n1e/f/ADf/AG2s y76oi7qd2Sb2VYeQLG5GNSy1jrhbS7DsbmO+1uw7vpNu3/YK7fUqp/SIOP8AUb0n4lx6jaci iq5uTe1ga+7IsbeyjPkuf6d+H9uy/R/nv5yv/QraajpD619A+z2ZD8o1V0msWerXbU4etIx3 Gq6uu307tv6K3Z6b1F31w+rbbHVHNaLGl7SzZZu31nZZjtb6e5+U3/uIz9Z/wnpbFmYn1HyK Xtsvz2Wva7CJc2mwF32K85u6x2RmZT3XZW7bbZu2f4T0loY/1adRk4t/2ncMXPzM/Zsjd9sG S30Z9T2+h9s/nP8ACbP5utJTpY3UsLLqx7sW0X1ZbfUoezUFgE7/AOT+7/XU/tuN6hr3iRYK Z7by3fsn5f56yOl/Vh3T8PFxTktt+zNa11oq2vlllmS11BNlnob/AFtl3896jFZHQ4Y6r7VY anOaRozcA2t9Eb9u3e7fv9T0/wDwX9KoRLN7lGI9viPqv9Dh9P8Az2UxxcFifroemv0r1/5r /9D0frH9Cd/ReR/Tf5n+0g9C/m7v6F9P/tDx/wBf/wCEXzSko/8AKfRT9VJL5VSUin6qSXyq kkp+qkl8qpJKf//ZADhCSU0EIQAAAAAAXQAAAAEBAAAADwBBAGQAbwBiAGUAIABQAGgAbwB0 AG8AcwBoAG8AcAAAABcAQQBkAG8AYgBlACAAUABoAG8AdABvAHMAaABvAHAAIABDAEMAIAAy ADAAMQA5AAAAAQA4QklNBAYAAAAAAAcACAAAAAEBAP/hFIBodHRwOi8vbnMuYWRvYmUuY29t L3hhcC8xLjAvADw/eHBhY2tldCBiZWdpbj0i77u/IiBpZD0iVzVNME1wQ2VoaUh6cmVTek5U Y3prYzlkIj8+IDx4OnhtcG1ldGEgeG1sbnM6eD0iYWRvYmU6bnM6bWV0YS8iIHg6eG1wdGs9 IkFkb2JlIFhNUCBDb3JlIDUuNi1jMTQ1IDc5LjE2MzQ5OSwgMjAxOC8wOC8xMy0xNjo0MDoy MiAgICAgICAgIj4gPHJkZjpSREYgeG1sbnM6cmRmPSJodHRwOi8vd3d3LnczLm9yZy8xOTk5 LzAyLzIyLXJkZi1zeW50YXgtbnMjIj4gPHJkZjpEZXNjcmlwdGlvbiByZGY6YWJvdXQ9IiIg eG1sbnM6eG1wPSJodHRwOi8vbnMuYWRvYmUuY29tL3hhcC8xLjAvIiB4bWxuczpkYz0iaHR0 cDovL3B1cmwub3JnL2RjL2VsZW1lbnRzLzEuMS8iIHhtbG5zOnhtcE1NPSJodHRwOi8vbnMu YWRvYmUuY29tL3hhcC8xLjAvbW0vIiB4bWxuczpzdEV2dD0iaHR0cDovL25zLmFkb2JlLmNv bS94YXAvMS4wL3NUeXBlL1Jlc291cmNlRXZlbnQjIiB4bWxuczpzdFJlZj0iaHR0cDovL25z LmFkb2JlLmNvbS94YXAvMS4wL3NUeXBlL1Jlc291cmNlUmVmIyIgeG1sbnM6cGhvdG9zaG9w PSJodHRwOi8vbnMuYWRvYmUuY29tL3Bob3Rvc2hvcC8xLjAvIiB4bXA6Q3JlYXRvclRvb2w9 IkFkb2JlIFBob3Rvc2hvcCBDQyAyMDE5IChNYWNpbnRvc2gpIiB4bXA6Q3JlYXRlRGF0ZT0i MjAxNy0wOC0wOVQxMTo1ODo0MCswMjowMCIgeG1wOk1ldGFkYXRhRGF0ZT0iMjAxOC0xMS0x M1QxMTozMDoxOCswMTowMCIgeG1wOk1vZGlmeURhdGU9IjIwMTgtMTEtMTNUMTE6MzA6MTgr MDE6MDAiIGRjOmZvcm1hdD0iaW1hZ2UvanBlZyIgeG1wTU06SW5zdGFuY2VJRD0ieG1wLmlp ZDo4ODQ5MzExNi05Yzg5LTRjNDQtYjdjMS0wNTU1NTg2NzBjOTciIHhtcE1NOkRvY3VtZW50 SUQ9ImFkb2JlOmRvY2lkOnBob3Rvc2hvcDoyYTgxODBlMi02YmRlLWU1NDYtOWNhZi03MTEy YjkzOTIxMGEiIHhtcE1NOk9yaWdpbmFsRG9jdW1lbnRJRD0ieG1wLmRpZDozYWMzMDRlNi01 ZDFiLTQ5Y2ItYTFhZi05MDhmNTAxMzJiNzgiIHBob3Rvc2hvcDpMZWdhY3lJUFRDRGlnZXN0 PSJDRENGRkE3REE4QzdCRTA5MDU3MDc2QUVBRjA1QzM0RSIgcGhvdG9zaG9wOkNvbG9yTW9k ZT0iMyIgcGhvdG9zaG9wOklDQ1Byb2ZpbGU9InNSR0IgSUVDNjE5NjYtMi4xIj4gPHhtcE1N Okhpc3Rvcnk+IDxyZGY6U2VxPiA8cmRmOmxpIHN0RXZ0OmFjdGlvbj0iY3JlYXRlZCIgc3RF dnQ6aW5zdGFuY2VJRD0ieG1wLmlpZDozYWMzMDRlNi01ZDFiLTQ5Y2ItYTFhZi05MDhmNTAx MzJiNzgiIHN0RXZ0OndoZW49IjIwMTctMDgtMDlUMTE6NTg6NDArMDI6MDAiIHN0RXZ0OnNv ZnR3YXJlQWdlbnQ9IkFkb2JlIFBob3Rvc2hvcCBDQyAyMDE3IChNYWNpbnRvc2gpIi8+IDxy ZGY6bGkgc3RFdnQ6YWN0aW9uPSJzYXZlZCIgc3RFdnQ6aW5zdGFuY2VJRD0ieG1wLmlpZDow ODc1NDUyOS0yOWY5LTQ3NTMtYTdjYi00Yjg0NjgzYjVhYWMiIHN0RXZ0OndoZW49IjIwMTct MDgtMDlUMTI6MDM6MzkrMDI6MDAiIHN0RXZ0OnNvZnR3YXJlQWdlbnQ9IkFkb2JlIFBob3Rv c2hvcCBDQyAyMDE3IChNYWNpbnRvc2gpIiBzdEV2dDpjaGFuZ2VkPSIvIi8+IDxyZGY6bGkg c3RFdnQ6YWN0aW9uPSJzYXZlZCIgc3RFdnQ6aW5zdGFuY2VJRD0ieG1wLmlpZDpkMzM4MDll Yi1mOGM4LTQzOGQtYTQ2Zi0xMWNjODQ4ODAzMWYiIHN0RXZ0OndoZW49IjIwMTgtMTEtMTNU MTE6MjE6MjIrMDE6MDAiIHN0RXZ0OnNvZnR3YXJlQWdlbnQ9IkFkb2JlIFBob3Rvc2hvcCBD QyAyMDE5IChNYWNpbnRvc2gpIiBzdEV2dDpjaGFuZ2VkPSIvIi8+IDxyZGY6bGkgc3RFdnQ6 YWN0aW9uPSJjb252ZXJ0ZWQiIHN0RXZ0OnBhcmFtZXRlcnM9ImZyb20gYXBwbGljYXRpb24v dm5kLmFkb2JlLnBob3Rvc2hvcCB0byBpbWFnZS9qcGVnIi8+IDxyZGY6bGkgc3RFdnQ6YWN0 aW9uPSJkZXJpdmVkIiBzdEV2dDpwYXJhbWV0ZXJzPSJjb252ZXJ0ZWQgZnJvbSBhcHBsaWNh dGlvbi92bmQuYWRvYmUucGhvdG9zaG9wIHRvIGltYWdlL2pwZWciLz4gPHJkZjpsaSBzdEV2 dDphY3Rpb249InNhdmVkIiBzdEV2dDppbnN0YW5jZUlEPSJ4bXAuaWlkOmVmYjA0MGI0LTM3 NzctNDBiMi1hZDY5LWZjMWU5YjkxOTZhYiIgc3RFdnQ6d2hlbj0iMjAxOC0xMS0xM1QxMToy MToyMiswMTowMCIgc3RFdnQ6c29mdHdhcmVBZ2VudD0iQWRvYmUgUGhvdG9zaG9wIENDIDIw MTkgKE1hY2ludG9zaCkiIHN0RXZ0OmNoYW5nZWQ9Ii8iLz4gPHJkZjpsaSBzdEV2dDphY3Rp b249InNhdmVkIiBzdEV2dDppbnN0YW5jZUlEPSJ4bXAuaWlkOjg4NDkzMTE2LTljODktNGM0 NC1iN2MxLTA1NTU1ODY3MGM5NyIgc3RFdnQ6d2hlbj0iMjAxOC0xMS0xM1QxMTozMDoxOCsw MTowMCIgc3RFdnQ6c29mdHdhcmVBZ2VudD0iQWRvYmUgUGhvdG9zaG9wIENDIDIwMTkgKE1h Y2ludG9zaCkiIHN0RXZ0OmNoYW5nZWQ9Ii8iLz4gPC9yZGY6U2VxPiA8L3htcE1NOkhpc3Rv cnk+IDx4bXBNTTpEZXJpdmVkRnJvbSBzdFJlZjppbnN0YW5jZUlEPSJ4bXAuaWlkOmQzMzgw OWViLWY4YzgtNDM4ZC1hNDZmLTExY2M4NDg4MDMxZiIgc3RSZWY6ZG9jdW1lbnRJRD0iYWRv YmU6ZG9jaWQ6cGhvdG9zaG9wOjViZjRiNGIxLTA5NGQtNzc0MC1hY2JjLWJkZjRiNjhjMzg0 MyIgc3RSZWY6b3JpZ2luYWxEb2N1bWVudElEPSJ4bXAuZGlkOjNhYzMwNGU2LTVkMWItNDlj Yi1hMWFmLTkwOGY1MDEzMmI3OCIvPiA8cGhvdG9zaG9wOkRvY3VtZW50QW5jZXN0b3JzPiA8 cmRmOkJhZz4gPHJkZjpsaT5hZG9iZTpkb2NpZDpwaG90b3Nob3A6MmExZGFjZjAtNmViYi1i ZjQ2LThjMTgtZTMzOGY4YzczNjQ0PC9yZGY6bGk+IDxyZGY6bGk+YWRvYmU6ZG9jaWQ6cGhv dG9zaG9wOjU1ZTMwMmQ4LWY1NTQtMTE3YS05YTBiLWUwZjRkYjk5OGU1OTwvcmRmOmxpPiA8 cmRmOmxpPmFkb2JlOmRvY2lkOnBob3Rvc2hvcDo2MDAwNGNiZC1iZTQ2LTExN2EtODNhMy1h NzYyZDdlZDYyMjk8L3JkZjpsaT4gPHJkZjpsaT5hZG9iZTpkb2NpZDpwaG90b3Nob3A6NzZj YmQ4ZTYtOTM0Mi0xMTdiLThkYjktZTA5ODg4OWY1YzBlPC9yZGY6bGk+IDxyZGY6bGk+eG1w LmRpZDpmOWQyYmFmYS0zNzViLTRhYjEtYmQ4NS02N2RhZjI5ZDA5MWI8L3JkZjpsaT4gPC9y ZGY6QmFnPiA8L3Bob3Rvc2hvcDpEb2N1bWVudEFuY2VzdG9ycz4gPC9yZGY6RGVzY3JpcHRp b24+IDwvcmRmOlJERj4gPC94OnhtcG1ldGE+ICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgPD94cGFja2V0IGVuZD0idyI/Pv/iDFhJQ0NfUFJP RklMRQABAQAADEhMaW5vAhAAAG1udHJSR0IgWFlaIAfOAAIACQAGADEAAGFjc3BNU0ZUAAAA AElFQyBzUkdCAAAAAAAAAAAAAAABAAD21gABAAAAANMtSFAgIAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAEWNwcnQAAAFQAAAAM2Rlc2MAAAGEAAAA bHd0cHQAAAHwAAAAFGJrcHQAAAIEAAAAFHJYWVoAAAIYAAAAFGdYWVoAAAIsAAAAFGJYWVoA AAJAAAAAFGRtbmQAAAJUAAAAcGRtZGQAAALEAAAAiHZ1ZWQAAANMAAAAhnZpZXcAAAPUAAAA JGx1bWkAAAP4AAAAFG1lYXMAAAQMAAAAJHRlY2gAAAQwAAAADHJUUkMAAAQ8AAAIDGdUUkMA AAQ8AAAIDGJUUkMAAAQ8AAAIDHRleHQAAAAAQ29weXJpZ2h0IChjKSAxOTk4IEhld2xldHQt UGFja2FyZCBDb21wYW55AABkZXNjAAAAAAAAABJzUkdCIElFQzYxOTY2LTIuMQAAAAAAAAAA AAAAEnNSR0IgSUVDNjE5NjYtMi4xAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAAABYWVogAAAAAAAA81EAAQAAAAEWzFhZWiAAAAAAAAAAAAAAAAAA AAAAWFlaIAAAAAAAAG+iAAA49QAAA5BYWVogAAAAAAAAYpkAALeFAAAY2lhZWiAAAAAAAAAk oAAAD4QAALbPZGVzYwAAAAAAAAAWSUVDIGh0dHA6Ly93d3cuaWVjLmNoAAAAAAAAAAAAAAAW SUVDIGh0dHA6Ly93d3cuaWVjLmNoAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAGRlc2MAAAAAAAAALklFQyA2MTk2Ni0yLjEgRGVmYXVsdCBSR0IgY29s b3VyIHNwYWNlIC0gc1JHQgAAAAAAAAAAAAAALklFQyA2MTk2Ni0yLjEgRGVmYXVsdCBSR0Ig Y29sb3VyIHNwYWNlIC0gc1JHQgAAAAAAAAAAAAAAAAAAAAAAAAAAAABkZXNjAAAAAAAAACxS ZWZlcmVuY2UgVmlld2luZyBDb25kaXRpb24gaW4gSUVDNjE5NjYtMi4xAAAAAAAAAAAAAAAs UmVmZXJlbmNlIFZpZXdpbmcgQ29uZGl0aW9uIGluIElFQzYxOTY2LTIuMQAAAAAAAAAAAAAA AAAAAAAAAAAAAAAAAAAAdmlldwAAAAAAE6T+ABRfLgAQzxQAA+3MAAQTCwADXJ4AAAABWFla IAAAAAAATAlWAFAAAABXH+dtZWFzAAAAAAAAAAEAAAAAAAAAAAAAAAAAAAAAAAACjwAAAAJz aWcgAAAAAENSVCBjdXJ2AAAAAAAABAAAAAAFAAoADwAUABkAHgAjACgALQAyADcAOwBAAEUA SgBPAFQAWQBeAGMAaABtAHIAdwB8AIEAhgCLAJAAlQCaAJ8ApACpAK4AsgC3ALwAwQDGAMsA 0ADVANsA4ADlAOsA8AD2APsBAQEHAQ0BEwEZAR8BJQErATIBOAE+AUUBTAFSAVkBYAFnAW4B dQF8AYMBiwGSAZoBoQGpAbEBuQHBAckB0QHZAeEB6QHyAfoCAwIMAhQCHQImAi8COAJBAksC VAJdAmcCcQJ6AoQCjgKYAqICrAK2AsECywLVAuAC6wL1AwADCwMWAyEDLQM4A0MDTwNaA2YD cgN+A4oDlgOiA64DugPHA9MD4APsA/kEBgQTBCAELQQ7BEgEVQRjBHEEfgSMBJoEqAS2BMQE 0wThBPAE/gUNBRwFKwU6BUkFWAVnBXcFhgWWBaYFtQXFBdUF5QX2BgYGFgYnBjcGSAZZBmoG ewaMBp0GrwbABtEG4wb1BwcHGQcrBz0HTwdhB3QHhgeZB6wHvwfSB+UH+AgLCB8IMghGCFoI bgiCCJYIqgi+CNII5wj7CRAJJQk6CU8JZAl5CY8JpAm6Cc8J5Qn7ChEKJwo9ClQKagqBCpgK rgrFCtwK8wsLCyILOQtRC2kLgAuYC7ALyAvhC/kMEgwqDEMMXAx1DI4MpwzADNkM8w0NDSYN QA1aDXQNjg2pDcMN3g34DhMOLg5JDmQOfw6bDrYO0g7uDwkPJQ9BD14Peg+WD7MPzw/sEAkQ JhBDEGEQfhCbELkQ1xD1ERMRMRFPEW0RjBGqEckR6BIHEiYSRRJkEoQSoxLDEuMTAxMjE0MT YxODE6QTxRPlFAYUJxRJFGoUixStFM4U8BUSFTQVVhV4FZsVvRXgFgMWJhZJFmwWjxayFtYW +hcdF0EXZReJF64X0hf3GBsYQBhlGIoYrxjVGPoZIBlFGWsZkRm3Gd0aBBoqGlEadxqeGsUa 7BsUGzsbYxuKG7Ib2hwCHCocUhx7HKMczBz1HR4dRx1wHZkdwx3sHhYeQB5qHpQevh7pHxMf Ph9pH5Qfvx/qIBUgQSBsIJggxCDwIRwhSCF1IaEhziH7IiciVSKCIq8i3SMKIzgjZiOUI8Ij 8CQfJE0kfCSrJNolCSU4JWgllyXHJfcmJyZXJocmtyboJxgnSSd6J6sn3CgNKD8ocSiiKNQp Bik4KWspnSnQKgIqNSpoKpsqzysCKzYraSudK9EsBSw5LG4soizXLQwtQS12Last4S4WLkwu gi63Lu4vJC9aL5Evxy/+MDUwbDCkMNsxEjFKMYIxujHyMioyYzKbMtQzDTNGM38zuDPxNCs0 ZTSeNNg1EzVNNYc1wjX9Njc2cjauNuk3JDdgN5w31zgUOFA4jDjIOQU5Qjl/Obw5+To2OnQ6 sjrvOy07azuqO+g8JzxlPKQ84z0iPWE9oT3gPiA+YD6gPuA/IT9hP6I/4kAjQGRApkDnQSlB akGsQe5CMEJyQrVC90M6Q31DwEQDREdEikTORRJFVUWaRd5GIkZnRqtG8Ec1R3tHwEgFSEtI kUjXSR1JY0mpSfBKN0p9SsRLDEtTS5pL4kwqTHJMuk0CTUpNk03cTiVObk63TwBPSU+TT91Q J1BxULtRBlFQUZtR5lIxUnxSx1MTU19TqlP2VEJUj1TbVShVdVXCVg9WXFapVvdXRFeSV+BY L1h9WMtZGllpWbhaB1pWWqZa9VtFW5Vb5Vw1XIZc1l0nXXhdyV4aXmxevV8PX2Ffs2AFYFdg qmD8YU9homH1YklinGLwY0Njl2PrZEBklGTpZT1lkmXnZj1mkmboZz1nk2fpaD9olmjsaUNp mmnxakhqn2r3a09rp2v/bFdsr20IbWBtuW4SbmtuxG8eb3hv0XArcIZw4HE6cZVx8HJLcqZz AXNdc7h0FHRwdMx1KHWFdeF2Pnabdvh3VnezeBF4bnjMeSp5iXnnekZ6pXsEe2N7wnwhfIF8 4X1BfaF+AX5ifsJ/I3+Ef+WAR4CogQqBa4HNgjCCkoL0g1eDuoQdhICE44VHhauGDoZyhteH O4efiASIaYjOiTOJmYn+imSKyoswi5aL/IxjjMqNMY2Yjf+OZo7OjzaPnpAGkG6Q1pE/kaiS EZJ6kuOTTZO2lCCUipT0lV+VyZY0lp+XCpd1l+CYTJi4mSSZkJn8mmia1ZtCm6+cHJyJnPed ZJ3SnkCerp8dn4uf+qBpoNihR6G2oiailqMGo3aj5qRWpMelOKWpphqmi6b9p26n4KhSqMSp N6mpqhyqj6sCq3Wr6axcrNCtRK24ri2uoa8Wr4uwALB1sOqxYLHWskuywrM4s660JbSctRO1 irYBtnm28Ldot+C4WbjRuUq5wro7urW7LrunvCG8m70VvY++Cr6Evv+/er/1wHDA7MFnwePC X8Lbw1jD1MRRxM7FS8XIxkbGw8dBx7/IPci8yTrJuco4yrfLNsu2zDXMtc01zbXONs62zzfP uNA50LrRPNG+0j/SwdNE08bUSdTL1U7V0dZV1tjXXNfg2GTY6Nls2fHadtr724DcBdyK3RDd lt4c3qLfKd+v4DbgveFE4cziU+Lb42Pj6+Rz5PzlhOYN5pbnH+ep6DLovOlG6dDqW+rl63Dr ++yG7RHtnO4o7rTvQO/M8Fjw5fFy8f/yjPMZ86f0NPTC9VD13vZt9vv3ivgZ+Kj5OPnH+lf6 5/t3/Af8mP0p/br+S/7c/23////uAA5BZG9iZQBkQAAAAAH/2wCEAAEBAQEBAQEBAQEBAQEB AQEBAQEBAQEBAQEBAQEBAQEBAQEBAQEBAQEBAQECAgICAgICAgICAgMDAwMDAwMDAwMBAQEB AQEBAQEBAQICAQICAwMDAwMDAwMDAwMDAwMDAwMDAwMDAwMDAwMDAwMDAwMDAwMDAwMDAwMD AwMDAwMDA//AABEIAEoA/QMBEQACEQEDEQH/3QAEACD/xAGiAAAABgIDAQAAAAAAAAAAAAAH CAYFBAkDCgIBAAsBAAAGAwEBAQAAAAAAAAAAAAYFBAMHAggBCQAKCxAAAgEDBAEDAwIDAwMC Bgl1AQIDBBEFEgYhBxMiAAgxFEEyIxUJUUIWYSQzF1JxgRhikSVDobHwJjRyChnB0TUn4VM2 gvGSokRUc0VGN0djKFVWVxqywtLi8mSDdJOEZaOzw9PjKThm83UqOTpISUpYWVpnaGlqdnd4 eXqFhoeIiYqUlZaXmJmapKWmp6ipqrS1tre4ubrExcbHyMnK1NXW19jZ2uTl5ufo6er09fb3 +Pn6EQACAQMCBAQDBQQEBAYGBW0BAgMRBCESBTEGACITQVEHMmEUcQhCgSORFVKhYhYzCbEk wdFDcvAX4YI0JZJTGGNE8aKyJjUZVDZFZCcKc4OTRnTC0uLyVWV1VjeEhaOzw9Pj8ykalKS0 xNTk9JWltcXV5fUoR1dmOHaGlqa2xtbm9md3h5ent8fX5/dIWGh4iJiouMjY6Pg5SVlpeYmZ qbnJ2en5KjpKWmp6ipqqusra6vr/2gAMAwEAAhEDEQA/AN/j37r3XvfuvdQ6/I0GLppa3JVt Jj6OBC81XW1ENLTxKASWkmndI0AAvyfaHcdz27Z7ObcN2v4LWwjFXkldY0UDJLO5CgfaelFr aXV9OltZW0ktwxoFRSzE+gCgk/s6AbOfJ/qPE01XUY/NT7rFDWpjqobYp0rIoayWF5oo/vq2 fH4+ZXSNrtFLIFtzbi+M3Mv3yfYnYbS+u9u5lfeltrgQSfu9BKqyspZV8WV4IGBAPdHK4FO6 lRWUdt9lOf76a3hvNsSwMsZkT6ltBKAgE6EWSQEEjDKpPlXPQf5P5o9f4anxtTkdr7wijyvn ko/BFhasNS08ngkmkmp8tJSrMJwVMAkaZBYuqgreJd5/vFPaPl+12i73flbfI4b4O0JVbSQN HGxRmLpdGMMHFDEHMqCjOqBl1Cyz+7lzXuE13DZ71t7PBpD1My0ZhqAAaIMRpzr0hTwUmhoI O1/k91JuihosimXr8LS1tRJQpUZ7F1NFSxZGFVkmx89dEKmhiqY4nVzqkCFGBDfj3LPJP3y/ Ynnfbdv3a35guLCxuJTCsl5byRRrcKAzQPMviQrIqsrmr6CrAhjwAT3z2U5+2O6uLN9viuZ4 0DlYJVdjGahZAh0uVJBGFrUEEdD5S1dLXQR1NHUQVVNKqvFPTSxzwyI6h1ZJI2ZWDIwIseQR 7yfs72z3C3iu7C6jmtXAKvGwdWBAIIZSQQQQRQ8D1Fc0E9tK8NxC0cykgqwKkEGhBBoeI6ke 1PTXXvfuvde9+691737r3Xvfuvde9+691737r3Xvfuvde9+691737r3Xvfuvde9+691737r3 Xvfuvde9+691737r3Xvfuvde9+691737r3Xvfuvde9+691737r3Xvfuvde9+691737r3X//Q 3+PfuvdBJ2525huqdvz189LNntwy0sk+G2pjpoVyeSCTwUslXJ5XAo8RRz1Uf3FSwKRhuLtY ewHz9z5YcjbTNdtbtd7sUJhtYyBJJQqpY1+CJGZfEkOFB8zjpyKMyMBWi+vVdu6fkxunJUke G7U2rsnetHnpIc8Np1dEBiMFhZWnSgios1QVDyVuUnCufKzSinCWa8hKpgtz37r7rzDtv9X/ AHT5Q2ffdqvHW4+imj/Qt4TqEYjnjbU07Zo5L+FTPeSFFWw7puvLl6u58u7rPZ36gjxI2oxH mD8jQVHA/Z17d2xuud77KpOwthnIbMwm1dvtkNx9cNj6jP5LE4v7x45t44ejpqhKnI4moZGN RUkFmSHVYAafeNXvV92Tk/3A5W27n/2huJ9k2HaNtZ7vYfBe8miTxCG3CzRZVaVJe7xpGOph EGUimjrJP2p95N5vtzflzf447vf7+cCG5eRIA7EUSCRyukZxCBQam0kVNeguq8pRYPZ+25Nk ZCjzO187PVZOp3PPQ0cy1e4IFFPPiY6Gr+9iwiY6geJwit5KxZRMxKqqphfzLu03t5yRynZc i3guOWr+eWe5vbq1jEjblEPDksjaXHjCxFtA0bqoYvfLKtyzsixpFONtYXG7cwbwvMtq9vvd qqxrbK7jRbsdSymRdBmMkgYVIpCVMYAJZm4VEVPvbasOS3Fk6Tb527XfwXC5SOimXE5dciZM hW4xcVhoC0eSx5HmlrI47NE6xzG4j96Xc19w+Qrbfecd/g2d9mu/orO6ELC1vBcaria1FrZx krc2/wDbyXkcelonSK4OoQnq0Zl5a36Wy2iyku1vIvGmiLjxYfDpGknizNmOT4FhZqh1ZoxQ v0qNqdjbh6SxmCzm3dz1e5463L5ClyEWMrJ320mPooqdqfDZfHZakR6jNzpI80cgNNLDTaQs rchZ59sffDmL2E2LlfmDlvnWXfIrq/ninFtI529IYVj8O0uoLqJWe8dWaaNx4Dx2+lUmejBS DfeTtp9yb3dNr3fZUsmjt42jMiL9QZHLapopImIWFSAjL+orSVLIuK2a9Od1bV7lwZyeDMtF kqRUGWwVYymsoWclUmRkJjqqSRlI1pcq3pYA2v2h9jPfrk3325aG98uO0G5RAfUWshHixE1A IIw8ZIpqX4T2uFNK4W+4ftvvvt1ugstzAkspCfCmSuh6ZINcqwGSDxGVqK0GT3OPUede9+69 1737r3Xvfuvde9+691737r3Xvfuvde9+691737r3Xvfuvde9+691737r3Xvfuvde9+691737 r3Xvfuvde9+691737r3Xvfuvde9+691737r3Xvfuvde9+691737r3X//0d87e+8sB17s/c++ d05CHF7d2lg8juDM5Coa0VNQYylkqp2NvU8jLHpRQCzuQoBJA9otwv7bbLG73G8lCWsEbOxP kqip/wA3zOOvda2yb/7I+R+Q3J27QSbi3XFvndOSSsingp8JjtlU+AkipNv7GoRW5IUmK27B hK2lrI2qJUmrauommmGpRbnv7kPvnP13Dv8At0MtzHcNIj8EWLw2HhwVZgqRhXDjUwLuzM3w ii2zcLq1Gma9CBk97NtCSg2NS0WAylTsyGrxeRzGawdFl6lc7VVT1mcx+KOSSSKDBYrIM0NO pjPkkWSa4EoAjfebw7GbXl+2tLSaWwVo3lmgSZvFdi0qR+ICFhjclUGk6m1yVAenSxO4liSC f9Xl0YHaO9ZMPunYfbw3FtvbkuYxsKbj2xn562kjyeMx8n938wcXj6HG11P/AHVz1BTa6aIi OOKdXEa6VU+xLtO5T7PzNyp7kQ7xZWU08IFzbTl0WWKM+BMI40jdfpZ411RoQqpIrBFooPVg XAKIWDq1VYGhDDINaghgfMdBp3nU1/xj3zWbQ6maro9rbsyH+k5q+spopcVn8RnhMNu4HBrU JOkGI2xRSyxtURhJpquU6xojjX3jt94XZIParmLcuRuSJriz2nc5huq3gSMmWKbWLSC38RXV rezR5EYkBpJHYNRUjHXS72qex97uVLHmnnIxz7pa262CqjEPE8QUzXExUqTLcsEYIxKLEo09 zselNs7eK937WyaVORpqLMbKgGYSsy7wY2ix9DlKunoq7DZOSjp1gLVtZ45aCoWJnlZJI24+ mJ99tHNHuPy/vWzc4bvbJuPLsZvbPc59FtYta3MsUFxZXLRRhI55JTFLat4bSuyyxEmPSVTb 9sY9uN7spLe0eTb9ybwTHFqkd3iRnSaMO2qiJqWdCwCgo4zxf4MVHtfZ+6clkZNu7niyVThc JS4ehyrZXGw1byzZD+OZY4+Wjq8fNSUlI9PRNqjeSSokFyqspLNh2dfbzkLnnfd2u9m323vb iysobS3ujd26Sl3uDfXZt2hlt3iiie3sjqjaSSeYVZI2Ri6W/ffOYNjsrNLyxeFJpnleIRSF QBH4EXiB0kDuyyTCjBQiHBYEJzbnZm6NhVW3t39cUVRjjsyvrJd07SoKivyVDncTmKqGeqra 3yCWpaiqI6daQs+o0ckcLhyGPuV/aj3j5n5Ln5Y5t9stultF2SeQ7ht0Mk08V1b3Miu80moM 5icItuS2r6ZkgcOQ5oY7zyXsnNEG7bBzhcpN+8Y0FtdSKkbwSwqVVEoQodSxloP7ZWkUrUDq 7jaO6MZvTbOD3Vh3L47O42lyNOGKGWD7iJWlpKgISqVdHMWilX+zIjD8e/oL5R5n2znPlnY+ atneu3X9skyA01LrUEo9K0eNqo4/C6keXXNbftlvOXd63PY9wWl3azNG3o2k0DLXirijKfNS D59KP2Iuijr3v3Xuve/de697917r3v3Xuve/de697917r3v3Xuve/de697917r3v3Xuve/de 697917r3v3Xuve/de697917r3v3Xuve/de6Su598bQ2XHSS7s3FidvpXySRURylXFTfdPCqv KIA51P41caiBYXF/r7APPfun7c+2MO33HuDzpt+zxXbMsJupli8UoAXCAmraQQWIFBUVOR0e bJyzzBzG1wmw7PcXbRAF/CQtpBwNVOFaGnrQ+nTng89h9zYynzOAyNNlcVV6zS19G/kpqgRS NE7Qy2AkVZEK3HFwfZ7yrzZy3zvsVlzNylvMG4cv3OrwriE6opArFGKNwYBlK1GKg0PSLc9r 3HZr2bbd1tHgv46ao3FGWoBFR5YIND07+xD0g6ReZ7F2Lt3M0m3c5urC4zO1/wBr9piqutjj rp/vZjT0gSC5e9TMNKXtqPuMuZvef2p5M5l27k3mrn/a7Dmq78LwbSWZVnk8Z/DhonH9R+1K 01HhjoR7byhzRu+3XG77ZsVzPtcWrXKiEougamq3DtXJ9PPpae5N6DnX/9LaS/mudi0lL1rt zpmrz1Vt7Fdgmsz26snQ4yTM1iYva2UwUmExSYyKvxprI8/m59DI0qJanJc6A3uCvfPeo7bY 7fY5LpoYbqryMqlzpjZNChQy11uaUqB21JpXrY6puzWBTB/wPCbbNXldr0m2MHVbWIpPDXbj g3KIqyrzNXjoZalIs5k8xO0MsSvKITCsQchB7ws5qaNrvbtqtmMliYUMGKGXxRUuVBNHdyVI qdJULU06WwqQtV4+fQ+7o3psKk3DHQ7l2zk92brxOHwWH3nmsXvKTC4nL7kxGPgossaampcd Us09MIUpJ6kSkVVTBJMAA49q+YpeX03BYt02iW83WKGKO4kS4MSSTRoFkooRjUABGfV3urPj V1eNjTtNF6UPYdbWVm7DmMclTkdr7kxGLymy56GinkpaXaZpVo8ft9UpklipqnbL08lDUR31 GaJpTzJch7nWxlud6e9t4nk2q5hje2KISq2+nRHCAtQrQaTE44llLn4q9KIpAFp5+frX1/Po 0Pc+4cbkfj98bMzvLZEu6a6rwOWwlIuTrcht98Su16iSjlnhqaERZOofPUlRTlopf2NFMsij Wb+yz7yPMG1WXtl7E3vNPKJ3K5ntp4I/FlktvAFoSrMrxUlY3KSRVR6xhYVdRqJPWXP3U4t2 vJucoNp5gO3iIxM+mNZjN4lQlUk7FERjch17yZCpOmnRcocnDkuu81D1lgBszM7U3LR7z3Vt +kyWQz0u9dsmlGHpq81FeDWy020K6pbz0SAx+Ot+4/UhIxjvX5e9wPavmLlzlzl0bfLZ3qX1 7apLLcveW2nwYrmN5AZAdulerRLVQs/jEdrdZWvZy2HN+2y847r+8NuvbN7S2neOOAWlxq8V k0p2BrtFGiY0asPhcGAItdU5qjz2PzmXyNLVQbbpsBVRbzxhWSnqfKzRJicdQy1EQVMnPm2g ehmKto0s5BVXBxy5L5fl5Z3nnL+tME7+38Wzytf0BjNzE7qtlFAzqVS9a+MBt3o3gukkhVow 4IP54sZ9vuNrsbKaNt6e7Q2klQygAEyyOFNTEIQ4mWo1VCgglT1Ay9biqbaW5p9jy7jxGVpp sLX52WuylKa5Nv008yMcTW4qnoJlgizE9O9WHBPjCkcBvYr5e3DYrXk7mWf23O72W7xy2kt4 1xcRNKLJHYE20ttHAwRbl4WuQwqUCEYDDpRYW9/Pv2zR8zpZ3Fk6zJAEjbQZ2UH9VJWkBYxL IIqH4tQOSvR/vgZ2A+4tmbg29ksqlTlaSrXcaUcv3D1QpMjLPjqzLNUun20lNlMnjWlKo5ZZ 2lZgAwv1/wDuI89y8x8h7psm47kJdzjkF0EOovolZopJi5GgrPLEXorEiQylgNQ6xO+9Fyqu 0cxbXu1pYlLKRDAXGkLqjCyJEFB1Boo5AtSKFAgBOk9avXzo/ng/OXv75S7o+Mn8v1slsva+ H37l+sNmS7B2pQbt7i7c3Dt+sqsVlszSVWWoslS7ew82QoqhqWGlp1dKSPzTzf6nPDrFnoOz 11/wpv2zFHupM98taxgBVfZR726r3JUkMDIY6jbH3dY8hv8AWLwX/Fvx7917o5OzflX/ADZd 8/ywfnlunujevZ3TXyR+Je9us9ybR3BJ1fjut9+ZXruOCOu3/jsvS1eB/g+4KJsVNUVH3UNK un7YDV9PfuvdDd/wnT+fnyE+U2d+TvV/yT7g3N27uXbOO2D2DsTJbqGHFdi9u1cuW23ufH0h xOMxiPRrlUoJTrRyrznkBgPfuvdA7/woL/ma/IvoD5F9T/H34ud1bm6kqtqdez757WrdojCv WZbLb2rvttoYLJNlsVlBHFi8Jh5q0LGELGuQm4C+/de6s+/kPfN7enzP+Gs79vbxrN793dNb 7zWxN/7jyq0Uea3Bjsgf7ybL3Fk46Cno6XzVuFyDUuqOGNWNCTy2r37r3VAP80T+Zj88B/Mz 7Z+Pnxc+SfYfW+0sXvrrnpLaOz9pJtqbHJvSux2Aw+byKfxHb+SqpKyo3JmZGlBlKr4rAAA+ /de6t6/m598fzRejav4pdG/BOm7C35nt09X7gl7W3ltbqSl7J3dXbh2pJtHDUWWrs7VY2qwO 3JMzNWVc8uqFPLIxZbKlvfuvdU2Vm4v+FOeHp5tz1I+XckFOj1ktJBtXq3K6Y4wZHA21RYqq rpQFB/aSFmP0t7917ocv5fH/AAoY+SmC7y2h0N88Mdht2bT3RvDH9dV/YybUGwOx+s905HJJ hKKTee36SOmxOUxMWZnSDIKaWkraQN5AXVGjPuvdH5/4UafLf5M/FbaXxUrfjd3Xu/p6q3vu zsqg3TVbRGFd87R4vbuFq8VFVnL4rKL46KoqHdPGEuWNyffuvdU/p/OW/m0fLvZPWXx4+He2 961XYuyeusNQ9q9idfbNoN9dp9g7opo3grt35PJ1WIbanXmFrY41ZU8CTSzeRzLp0ovuvdCz /NP/AJkPzv8Aip869g7P2r352NtnaG0+pvjDvHfHVkY2yuNzuclwGNyvY+Ly5qcHV1KVO6qi mqIKpo5giySEpYD37r3TZvD+ab/Ot/mF5HdO4/g/1Nv/AK66Z27X1FJAnS+w6DdOUiMOpkpM 72hu+gkocxuLxR3lpMVDCEa6iM/n3Xugj+IX/Cgv5tfGnsLObd+WlRnfkhseij3Bhc7tPdeM wmy+2dnbwxi1MNLHTbhpsRjkRKbMwCnyNHkaSZo4S7RFZEAb3Xulpu3+Y5/Pk+bWCz3dPx12 B2f1z0ViWyFZQf6AusaSsxaY/HrJNULFvTd2OyOf31U0kEZ8746PxagdMS8D37r3Rlv5Ln87 P5K9kfJXaHxR+XW7qftPB9tvkMJ112PksLjcFvfa2/qDH1eTodu7gkw1Jj6DPYbckFBNTo0t PHV0taIxrZHZR7r3W4zlsrjsHjMhmMtVw0GMxVHUV+QrKhwkNLSUsTTTzSOeAscaE/1P49lu 8bvtuwbTuW+bxeJb7TZwPNNK5oqRxqWdifQKCfn5dKrGxu9zvbTbrC3aW9nkVI0UVLOxAUD7 Seqin3HkPln3tHC9auJ2bSOViepqYaYYjZtHUqGMQmkVWze4JSOBdg8g/sxe+C+8De/v1/eb gbcr02Pt7btpj8R1j+k2mKQV0ayAbu+Y1NKkNIPwQinQBdps/YT2ueRIPH5ikFSFUt4t26+d AaQwAeeCF9X6t2xWNx+GxtBicVSw0WNx1LBRUFJTqFhgpaeNYoI4wvGlUUc/U/U8++8ew7Jt HLOy7Vy9sFjHa7JZW8cMEUYokcUahUVR6BQM8TxJJPXP++vbvcby6v7+dpL2aRndm4szGrE/ mekh2d2Hg+rdlZvemflC0eKpj9vTBlWbJ5Ke8ePxdMDy09bUkKLD0rdjwD7B3u17mbD7Qcgc w8+8wv8A4pZwnw4waPPO3bBbxjzeWSi+ekamIop6P+S+Ut0545l2zlvaY63E79zeUca5kkb0 VFqfmaAZI6pax2/83vTuDbO6dySO2T3LvfbeRBYOsP2s2dpKelhoQ45x9JFF4YdPptH/AFv7 +dew3nmv3F+8Dyh7h86M777u3M233LMQQojN7EsSQ1/0CJE8KKmNKY8+ukV5yrtnLnt7vWx7 MoFlZbZcR4oTqEDsxen43J1vXNW+zq+T39OHXLTr/9PYV/mtx/3h7e2FtepyeAwVPjOucfua kymbvQwTVNVvDOYeox1VlYYKid08SpPBAy6FMcrggn3in94Rjc7nt1k1zDCiWYkDydoqZXQq WAJ4UIHDDHj1eMVanQEbY3turYmUxG0NpZxqTB7Go8VQUyrTUUlJnapVXMZXLNLU0slTJjs7 X1srRgOoFKy6QvvGTfebd95ZvIdv2Tc2itbBUC9qlZGH6jyElSxSV2JABA0EUp0YxRrIO4ce gX7K6uxu2uzstiMZvvYO28dmJ8XnsVht07hnx2W27j930tNmKbH5mCShmCfwo15TWHbyQxq/ BJANL7YbS/uI7q03OztobpY5RFNJpeAzKrmOQUNChb1ymljxp0yzFHZHPDz9elPvHtrdWzdy 1PXWytybm2ptXrkybRoKCkrKjFvla6jlMmd3ZkYI29dZurKSSVaG5C0bQKDYXOuYd43jaL59 h2Pcbm02vb6wIiMULsDWSdwOLTuS4P8AAUAwOtppOWAJPRwqTNZfun4k7Jym5dz4s5jYXbmd 2vLm9yVskE9Zjq/b0Nbj462sEVRLV1wSZEVmAJiiW5uOY6+8Vb3vO/sDyFum78xWdve7VzNN bGa8lMeuOa1LIpcK5aTC0qBVUycZya+6/vp2jm3ma0SyuJxcberaIVDsDHKuSCygChI+09B3 QY+XrHZWe3ngc5jsluvcGQOwsTm9sVslVHs3F1FIMpn6uesMUJo8zuOljSkpRp/4DioINz7x W2GaT2z5J3znPZOYLW85o3C5/dcFxYyGRdvhaMT3UjyaV8O4vEVYIMH9IXBBr1mBd3Kc6cyb Vy7uu2TQ7FaRfWyw3KBTdyK/hwIq1IaG3YmWTP8AaeEKU6beud87rzWeqNn7u3Nlcrt7c9DU RPUZaqepptuZPFQz5PFbnlbTqho8a0ciVjG4FHPKfqF9l+3b7vvulY7p7dc2cxXFxt+4Ql4Z Z3LpZ3lqrzW92x4rElJIrk5At5ZG+JR0r5t5a2Latth5i2HZoIN0spVOmJQrTxSlY5bYeryV Vohj9VEHAnpTiV9n4ncG8cXn8DlsnT0FLicauFqFyy487imFPPlshTVVKtM9DHRxPFH5FdGm lS49gzkbbJuRNu5g5x2zmHbbzeIIEtYVtZPqPBa8fRJPOkkYQxKitGusMrSyJ2jB6K2VeYL3 auXr7aruCyeV5ZPGUxGT6calijZW1ByxDNpIIRWz0Y/4H5qu3P25nc1ka6gFVketq/BZDHwx QY+asjwm46OWiylLjaKniokpUiyrwyyKFvML2JJPvo19xrc9x3zn7c9+3G7gF5NsLW8kaqsR kFtdI0cyQxosYQCUo7KBVxUgkk9Qp96HbLXaOQ9v2u0tpTbR7usyOS0gQz27h42kdi5YmMMq mtExUAAdaYndu0flj/KI/mIbn7Iwu2sjtzP7P7O7A3R1JvvPbVrc11x2P1/v2pzSgRV4CY+u ir9uZ6Shr6ZKmGto6lG/QwVj1L6wK6PJQ/8ACn753QVlPJWdZfGivgWVfJQPtff9F92oN2hS pj380sTlQSCoYj+h9+691sTfy+v5i/Wv84P44d8dZbj2OnW/Z+P2ZlthdsdeHJDO4Ss272Pt /L4TH7u2lk56ekqazAZBvuIXiqIlqKKpi0OXDJI3uvdatv8AIq35kPi5/NYxHV+86x8NHnaH un4+7zFU4giTK7VSuy1A9b5igREzmxI2Ja2lZC3v3XulV8XOr6T+cj/OI703hv2lq8p0/Xy9 t77zut5pI8fsTDY2brHpnHJKrKUanqpcbVQqCoZqQ/gke/de6X/8hvt7O/Cz+ZT2t8VO0qo4 Wl7Ept+dR7mpKuTwUtN2r0rWZvNbcrtMllEmZxONydNAePJ91AAeR7917oEP5WWzan5s/wA6 Sh7RysJyeDxvcfcPyfz08yGpgWh2/msrPswTnSUMc2cyWLjjB44H9PfuvdbIP817+eVtv4B7 7i6A6k66x3bvfa4LH7j3U+5MtVYjr7rfH5qI1GDpM42LU5rcO5MtRD7paCmelSCleOSWoUyI h917qlzG/wA+n+cVuylj3htP45bezGzq4mqoK3bvxm7bzu256Thv8i3BT5mqFdT6b2lSZwRy CffuvdUefI3vHfvyT+We4+8u0tm4rr7sXsLsnYWT3ftHCYPNbZx+KzFHXbaxUrrhNxT1Ocoa qujoVnm+5keR5ZGa9iPfuvdbOH/Cpn/mWfwW/wDDn7K/94zbfv3XurDP+E3+2tu4r+Whs7cG LwWIx2e3Z2b2lVbpzVHQU8GV3FU4vdVZi8ZNmq9EFTkHxuNhSnpxIzLDEulAATf3Xutar/hR 6zL/ADNOxWUlWXpnqRlYGzKw21WEFT+CD7917re4+I+09r7I+Lvx823s3buG2tt6h6c65ko8 JgcfTYzG00lbtLE1tZNHS0qRx+esraiSaaQgvLK7O5LMSfde6+eP/NKw+Km/m9fJTDzY+lkx OX+S+x4spjmiApMhFnY9jjMxVMS6Q6ZQVUvn/wBX5GJ5J9+6919KPbG38FtXbuD21tjDYzbu 3cFi6HF4TBYWigxuJxGMoqeOCjx+OoKRIqeko6aFAqRooVVHv3Xuvm2dCxRYn+dlsGnxcaY+ npP5jFZTUkFGoghpqc905WD7eCOOyxQ+GQppAACm3v3Xut2n+Y93VX7ax21eoMPNLAd10sm4 9zmnJ8tZiKWt+zxGHAQhzHXZOGSSRP8AdnhRfoTfnD/eAe4u5220cu+0+0zPHBuaG6vdPGSB JNEFvjOmSZWd1/F4arwJrmt90P27td3vN85/v41b6Fxb2+rgkrJrllziqRkKp8tbHiB0FGI/ l+9rZXr+m3NLu3BY7dtXjY8tSbKmpasCNpIBU0+Nq9wR1QipsmwIU2geGOVrF7AsIP2z+705 83HkmDmCXmeyg5okgEyWBR6Cq6lie6D0WbyxGUVsF6At0PL/AO9byPY81S7NHsNzNsMcxia8 DLmjaWkSArVo+Jy4dl4LWg6XXwF+QO7q7d+b6I3zWZCu+wx2TyG22y8r1GTwdbt+qSlz23Z6 mVnlko0VzJEHY+FonVTpYASv9x73Y5wh5h3j2W50v57mCCCWWzM7M81q9uwS4tWdiWMYrqQM T4ZR1U6SB0F/vS+1mwW+wbZ7n8sW0UXiyxpceEAscyTqWguAooA5ppYgDWGUnI6TPbm+M18y PkViOlev6uduttlZCpOcztGGaikGPkFPubc7yreJ0iJOPxhJs8shZf1mwY95Ny3f73Xvrsvt Pyncv/rc7NOxnmSuhihC3d4SMHSP8VtK4LsWX4zQ45B5d277vXtJuHuLzVbqOcNyiXwYXw41 gtb24ByCf7e4oMKAD8I6QnyRhx20PmB11s/CUqUOHww6WxOKoogAkFFBkaaCCNQANTWF2P1Z rk8n2Avf/lDaeXvvVe32x7HZrb7Pt45fggjUYSOKVVVR68MniSSTknoUe0E95v8A7Ac3b/uU 5k3G5/e8sjniztGzE/5h5CgGB1eJf/in0Pvs31zV6//U2Sv5rnXeZ3BujqXOYfG01TPV4LN4 uOY1FHQS1UuAyMORkxdRXZGspKAK9HnpJqWJ2VpJYnVCzMF94x/eF2S63B9luLWFWkMTqMha lCG0lmIXIclQaEkECpNOrIaMM9FFwmTwe1cxs7b+7NsS5jO7WwW2Kbd0hzE1JO9Qv+5E4V6e GGopKk4fE1ENJr1/vGOxa1j7xN5jbbtrvtps952dri6tYYBcfqlCaHX4RABVvDjKx1r3UpWm ejOGpDFWpWtOo3yn6oye9qCv3th1kzG6sEs2YSopqZ2/vlszJBsggokjU662hglEsEIJK2lh HqAHsNveXW18ybjDucpmsb2TxUm0nQ3i90Ug8grqQrAfAw0n4T1eeMSRBox+oo4eo8/zHReN 8bl2hjanZ+E31i9x5zsjB7F2/jt9ZPBZ7H4mmhrliM+AwWThrMbXSVm5NubWmpKSvqA63kQR MuuJiZU5gj22N9vt92tZ5d7htY0naORUGoCscbBlYtJHEUR2rxAWlV6LkcipHD06sh6omwW7 vhlvF9mo0O0tubl2FlUxdfNSVO4sNu4SVeO3zHuGrpYovv6iveWmrKSpEaRPQTxRKoMLj2Bf vIbZZ7n90nmi62iOm3bZuu23IRipljlMhhufEIA1FvEV1agHhsqgVUjqd/u5bl9D7pWEZb/c m1uIvkTo1gftX9vSK2pT0uP21vqvz1PLWbSyVDR4I4qKVaaqyu6zUpXYSoxlQ8ciUtTt+COe omn0PaGTxEfui3Or263G22rlP3A3XmSJ5uS7mCKz8BWCSXO5axNatA7BhHJZxrLNLLpYCJ/B I/WFM5d+kmvN55YtdqkEe/QyvP4hBZIrXSUmWRQQXWdiiIlQda6wezLVh8Xg8xhN9YvauJym P3VPt0VOPFZko8rJmMDQVaVW7cHQRwUNI8GTqsaiTLp1manppoQP3ORTyhPsO+bHz1tXKW1X dtza23CSESTrcNcWsLh7+2hVIoys7wBZQV1GSGGaED9TKvcrvctu3Pli+3y/gl2JbvTJpjMQ imkQrazOWdgY1kJQ1oEkkjkr24SFCuR2ht3P5jc2GrRh6rDNtugxmWp8hi4NyZDJzRz0kEc7 JBUGHDQ0rVkkkJ1J40W41i7XKWzXfL1rzJzFv+xzrtN1a/TJBPHLCl5POysoViEbTAIzcs8Z qpVRUaxU7vHtN/3Tadu2bcYzuEdx9Q8kTJI1ukYKuxWrLWZmESq+DqY07ej2/wAurAfxrcm6 uy6ego8ViqXZGL2ZBjKWrmnFNlZNxZTKZGpSKrlnrVgr6aCB9cjnXIGC8LYdLvuOcuLc7hvf O8FlFbWCbVFZLEjMwSX6iWWRgHLSaZFCNqZjqbUFwvWLv3td1O37RsfJ011JNfPuUl20jKBq jFvHFGpKhVqjFxpAwNNcnqvfvr/hQ71D0V8oPkB8Yu/Pi7n914vprs7KbLxO8NlZrbO5YczQ UdJQVMNdk9pbqgxsuPyLLW6XWKolQ6bj62HRvrBjohv8w/8Anf8AwQ+TfxL7Q6L6p+Lu6Z98 9iYaLE4XcW+djbE2tiev8gKylqU3dSZHGV+Uy02YxUUL/bLSiFmlYBpBGXB917pQf8JY+mOw B2J8mPkNUYnIUPV1TsfbHVGGzlVBNT47dO74dwS7ky0WHlkCxZKHbtBHEtRLFrSKaqWO4a49 +691Vj/Oi6xznxg/ml975bZ9bW7XO/a7C927PzGHklx9TSN2Ht/7bcpx86KhiZc7DkIZCp51 sPz7917rYe/4TF/GROvfix2T8lsxjlgz3f8Avh8FteplgKzr1v1lJU4enlpZWAJo8zvCpyTt bhxSRt7917qm/wD4UKdF5z40fzFcf37sOortsUfyC2tiuysNnsNK9FUYrs7ZyxbU3g9JVRBD T5Cro4sfWsQdTGocn6n37r3Vg/8Awle6IEGH+UfyXr6Qr/EsjtfpDadUY7I9Jh4v7671EMn1 0vkMliY2C8aoLH6e/de6qg/n79M9gdP/AMyns/sXduDqptk90vsrsLrrcGRpp325uOkw+2Nv YHO7ZNcw+2etweVwLwVVIGEq0k0T6dEgPv3XurlsH/wqT+OeB60wVJS/FDtLH7yxm3qGhOz8 FmNkYzr/AB1fR0cdMKLD5w1P3FPt6GSO0A+wEiQ6V0XHv3XutV35L/JXN/L/AOXe6PkluTbO I2Xl+1OytiZk7WwtZVV+Pw1Bj8ltzCYumhyFZFTz5GX7HGoZqjxxpLOXKKosPfuvdbNP/Cpo gdZ/BYEgH+8/ZXB/8Mzbfv3XurJ/+E53/brjq/8A8SL3D/73GQ9+691rHf8ACj8gfzMuxrkD /jDHUv5/7Nqs9+691vtfG/8A7J46G/8AEM9X/wDvE4P37r3Xzt/5oxH/AA8Z8g+R/wBlN9Zf 9CbA9+6919Kqn/4Dwf8ALGL/AKEX37r3XzWOlSP+HuNk8j/t49V/n/v92T9+691uU/zQuqNz T1Wx+7cBTVFTi9v487Y3PLSwvO2DaHJtl9v5qrRNRXHSVdTLBLIQEjfxhiA9/fOb79Pt5u9x ccre6G12zSWdnD9LcsoLGCkplt5nA/0MuzIzcFOipo1es+/uXc8bNFDzJ7cbpMiXt1L9Tbhi B4wMYiniUn/RAqq6qMsNZAOnpT7f/mj9d0/WlLPnNn7pk7LosRFSyYWhgpW23ksvBTJEtZFn mq1ajxdVOvkZWhaWJSVAYgEn20ffq5Ni5Jgk3Pl2+bneK3CmFFQ20kqqBrE5fsiZu4goXUYA YivRNuv3K+bZecZ4tt3+yHJ0k5YTOzC4jiZq6DBp75FB0ghwrEVJUYBcPhn1L2d25le9O56d JMZU5zZHYu3dqZeSSfF0OU7C39HO9VJjaxUZ0x+DEtpJ4w3jeVQLsp9wt92z25509wdw91Pc yCIwXN3te429rMS0SS7huAYuY3pURw1ozrXSzgDIPUufeE565M5DsvbP28lcTQ225WE9zEAJ XjsbIqFEi1oZJqVVGI1BScAjqPtz4B/MraDTybW3JtvbU1ZDDBWy4DsPL4mSrihbyRRVMlBj oJJ4o5CWCsSA3P19sbL90X7xvLhlfYN4sbGSRVVzb7hNCXC5AYxxqWANSAcA549Pbv8Aej+7 3vwiXe9ovLyONiyCexilCk4JUPIwBIwSPLHRXu0+ue49md0Yjr7f25P4n2hXT7Shx+ZO5shl /BLnKqODbhOeq0Stg+zqWBuBeG119wZz3yN7h8t+5m38o827v4/PUj2gjmN1JNRpnC2/670k XQ2eHZxHU08lc3e3/MXt3f8ANXK20eByXEt0ZIvp0iqIVJn/AEV7G1qKUr38D1cL/oL+S3+y ff6LP74N/ps/vP8AxH+8n9+cz/xbP7xffeP+9Hi/iVv4b6fHpt/Z+nvoz/rU+9f/AAOn9Qv6 xf8AMTPrvE+p+un/ALL6jxKfVU8X+zxppTy4dYBf65fs7/r+/wBdf6v/APMOfo/D+n+ih/tP A0V+mr4f9pnVWvnx6//V3TPmj0Q/yE+P+9Nk4yKA7tpKaPcWy5ZvGoO4cJKldDjnlcWjp85B C9HISQoEwY/p9gT3I5VPN/KW57VCo+uC64Sf9+J3Ba+QcAoftr5deHEda9m3o8Ri9sbJ/wBJ uUz+zN1lc7haOOg2/Q7jEm2tu5CLHUFXmof45jKjz0lfJV0dO1O0okgpACLgXwD3rZbIbftn 9YLq4tNxpJGpSFZqwxsFRpB4iGqtrRdJaqoKjpbDIRUqAR9vRs6zuPcm3c+mH21uPKYjauJx uHg2dTUddehqduR0UT4rOxRWeB6nOsz1k7AXFRLIjepCAHOad95k2zc/oNj3i5tdlgiiW1RH /TMCqDHKFNVLTEmVzSviMynhQLYlRkDMoLVz6/6h0WD5GfHeHc9fS94bDzWztr0+/qXIbh3t tHdWeXAtS7gpaySmzu8NsCaCoFTtvcdaj1DpdfDVGbRdB6RVDKu+Q7ZvF/uNpZ3e4Rs+ieUR M0qMUlkQEGsMzgujVFGLp+EdIZ7fSWaKpUHNPIdGJ+DGYo9h5mt6Ax9ficnuTvvpnO9g0++8 HXU+awGS3PgpqiXZm2NnZEiTGZSlwWMx+QWvmWMu+Unkh4WmTUPRs15LyTzz7ebHcAcx73y9 cz29xHpkU3cIL2kUJcNG+jQ6uSpJlcj8A6OuT9yttp5q5e3G+jD2Md1H4oJIBjY6XqVIPwkn BHDoV4pMt2LtnN43dmVgp6vZ8sOdxe5M346DH48ZGaLE5XblfJQUfEmVkRJqQCJ5fLTuv6Sb chNk37fvdzk7mbaOet/hhv8Al1kvbfcLzTDBB48i2tzt9wYYviuWCS2oETyeLBIuEJI6JSxW HJ+9bZebDYM8G4qYZLeGrySeGpliuIw78IgWSXvVdMin4hlPUeLp9l7e3Du3Bbkxma3FDJRb dpK7APk4m2dBnYKoVmfefIY/GyCvroKZ6Gjlg1rA0sjFlcxgrNjSz5A5U5m545b5wsty5nje GxilsTcL+7EvFk8S9Lzw27eNMkbWds8QcRNJI7Msnggms9/NzLu207BumzzW2zsHuHScRn6t oCuiDTG8g0IzCaVXoXCoACuvoJd1ZDX11n6vezbn3VQVGWwz7boqfOTHMS52hmeCokoqzIxZ Vosa9FkhTVAiTXLLNEqHUvB3yFzNJue039vztdbjudjuF1Etin1LtO15b1LtG8wmKw+FJ4Mu hayPJGqHUpKjfarMLzXtcPLf0VjdJBL9Q5hXwhBIAVDpGYqyB4/ETW2lVR2bBobvPjB1LTdR dV4nEikkocpnvFuPL0U92qcVLXUdKtFt+aZwJZ3wGPjjp5HbmSoEknBc++83sd7e2/tzyFt2 2JbmK8uaXEqNloi6LogJOWMCBY2J+Jw7/i65p+8/PU3PvO9/f+OJLK2rbxOvwyhHYvOAMKJ3 LSKB8KFF/COgv7z/AJb3wU+Seer9191fF7qbe27Mq5mym65dvjCbnylQUWP7rJ5/bc+HyuSq /GijyzyyPZQL2AHuYeon6L1tz+R5/Ky2xlI8vRfEXY9fUwSrNDDnszvTP46N0YMt8Zlty1VB OgI/TLG6n8j37r3Vnm0tnbT2DtzEbP2PtrA7P2ngKRKHCba2ziaHB4LE0cdylNj8XjYKaipI QzE2RBdiSeST7917oq3yO/l6fDH5c7twu/Pkd0Ds7tTd+3tvjauHz+dnz9LXUW3lyFVlExat hsxjI5YI6+tlkUyK7qZCAbce/de6MZ1X1X170h13tHqbqjamM2P11sTDwYHaO1MMsy43CYmn Z3ipKb7maoqXHklZ2eSR5HdizMSSffuvdBH8lfhn8YPmFjtp4r5LdO7Y7bodjZDI5TaUW4ny 0D4Kuy9LBR5OajqMNksZUaa6npo1kR3ZDoU6bgEe690rvj98b+j/AIrdeQ9U/HzrnB9X9ewZ jLbgTbOAavkozms7OtTlslJNk6yvrZqmtlRSxeVgAoCgKAPfuvdKLtfpnqXvXadTsXuXrfZf Z+0Kp/LJt/e+3cZuLGpUBWRaulhyVPP9lWojELPCY5lBNmF/fuvdEexX8nL+WHhs1Dn6D4Yd OjI0861EX3ePzGSx6yo2tT/BslmavDsgb+wYCn+Hv3Xuha7i/lz/AAb78k2TN2r8Yep9xzdc wU1Jsiak26m15dvUFHUQ1dFjKKTaM2CZsVRVUCyQUsmunhe5RFLNf3Xulb8kfhP8Wfl5j9mY r5I9N7b7Xx/XtRkKrZlLuCpzlPHgajK0lPQ5CWkOHyuNZ2qqSljRvIXACi1j7917oQOhfj10 z8YOuMd1H0JsLE9a9cYnIZbKY7auFmyM9BSZDOVj5DK1McmVrchWF62slaRgZSATwAOPfuvd F379/lk/BH5Rdh5Dtfv344bL7M7DyuJxeCyG6M5W7nhr6nEYWmekxdC6YvPY+kWKjppCilYw xB5JPPv3Xujr7fwOI2tgsLtnb9FHjMFt3EY3BYXGwtI0OPxGIo4aDG0ULTPJM0VLRwIil2Zi F5JPv3XuiI9j/wAqv+Xz252tne7+x/jFsbdXa25txY7dmd3nX1+7I8nktxYoUIx+Umjo9xU1 Cs9L/DYNISJU/bF1PN/de6sEChVVV4VQAAPwALAc3/Hv3Xuq+8R/Kr/l9YHt+i78w/xi2NQd wY/f7dpUe+oa7dhysG/3zMm4G3KsUm4noDWnMStPpMJi1H9FuPfuvdH8q6Olr6aooq6mp6yi q4ZKero6uGKopaqnmQxywVEEyPFNDLGxVlYFWBsR7ZuLe3u4Jra6gSW2kUq6OAyspFCrKagg jBBBBHHp2Cea1miubaZo7hGDKykqysDUMrChBByCDUHI6LDU/Cf4tVWa/jsvTm2RWGY1DQRN koMW8zOXLnEQV8eNA1H9AiEf4029wbN92T2Jn3I7rJ7cWP1BbUVBkWInjXwVkEf5adPy6meH 7xnvXBt42xPcC98ALpDHw2kApSnilDJ+erV8+jLYrEYvBY6jw+Fx1FicTjoEpqDG42lgoqGi p4/0QUtLTpHDDEv9FAHubbCwsdrs7bbtts4rewhUKkcaqiIo4BVUBVA9AOodvr693O7uNw3G 7knvpWLPJIxd3Y8SzMSSfmT04+1fSXoJtydE9P7w3bTb83P17trN7xo5MZLS7ir6LzZOCTCy LLinjn1jSaCRQ0fHBHuP969q/brmLmGDmvfOT7K65jjMRW4kSsqmE1iIav4DlfToc7P7mc/b BsU3LOzc13lty/IJA0CPSMiYUlBWn4xhvXoWfcgdAbr/1t/j37r3VKfz0+EMNHNu3vfrbHZ2 pw8qy7l3jsXZ2Jiy2dx+5Jqilhy279sYuWopIqjCZLHB6nM0EbiX7inWpp1LPKvvGf3b9qoL hdw5l2qCRoTWSWGJQzrISA0sakgFGWrSoM6l1qCSw6srFTUdVqP3Dh8DtLYFDgqPbm9tswYn J00+5t07fJyA3NW5Sor8ntuMPWQ5LAUmEp2iNPTGXTO0s1Qlw5C4zbsLa12zZ7e1sra8sEic GWeKreI7lniB1B41QU0pqoxLuOOFKSjJLEHoUs7ubOdiYzB9kbajoqbFVtLDtGbBVGWxeEp9 u5XamPo6afE7YjzNfRjJ7blo6iKoiaJpXglnkjnYycsX71sF7zIlnzBYwItu0a25iLJGsLW6 qvhwB2XVDpIdSCxVmZXOrJeWfTjV8/t6QtXuhfj/AA9b7lodu4yv7KxG8m7i2ZhavOZDHYLZ eMrBPishDj8ntetEFZJ2RUURmr8cZZsbAKZJGRamaUAYbLvN/wAkWnL67jb/AFO7xT/VQVcj 6eM9lC8bfqfUEEvHVlQKK97EBNKqsSyAD/L1aTl22l3709i+5OkJMdTbYy2elzvYW1ZK2OHI 7Y3fkKVI8pTbkR3c0km3q3yJAgC0zU1SJohZtRxm+9n93s7nyO/uJ7MQ2ltyXcb1JuO62vi+ Hou7pArXFwzk6EtJNaxxrSGOK5LogOpmyu9lvdK3utwh2vmaaWXdlslt7Z9OohIjXwlA4tKN JZjV2MYBJFAAlgxuU2lh2r8M1HnMjuj7rDrPiEi3Bi8fRY6po6mqpamKSnlpZczUVSwyRa0K wQjWh1OCuDW0w797e8vDdOX5rfdN33wy2ge0C39pBFbyRSSRSKY2ja8kk8F49SkQQ0ljbXKC mQ8t3Y8wbgLXcBJa2ljplIlJgld5FdVZSGDCFV1q2k1du1hRaE6vx9+P1fmJNt7+7cw4nr9u fcy7Vx2RUJIz1FaMjR1mTxaxR04XD1DO1IHF9b6ioEaFuqP3Svu1cwb03J/uv74bNIN52nxW 22GaqSfqyeKklxAqqoFszSfThskyaigWOMtjX7se69ttybzynyDuGm0vNIupI8iip4bJFJUm sq0EpHkKaiWYA/fvp91it1737r3Xvfuvde9+691737r3Xvfuvde9+691737r3Xvfuvde9+69 1737r3Xvfuvde9+691737r3Xvfuvde9+691737r3Xvfuvde9+691737r3Xvfuvde9+691737 r3Xvfuvdf//X3+PfuvdePv3XuiC99fAXrzsuk3RlerK6j6S3/u3JY3L7jzuE2xhs7gd1VeJS t+0g3BtXMQzYyOGonrDJUS0Ipaid1DStIRzGHM/tfs+8x3020Mu37ncMrSOkaOkhXVQPG4K0 JarFNLE5JPXuHVU3bn8vH5ZHCUC5zGYLd0+ymOA2tWdP0+MONy22K6etylXVVvXuTqdqRbdy 9LknvLNRPUNkDMPIgMQdoQ372b5rmghF1FFM1t2RG2C6WjJLEmBjGI3DHLIT4lcior1vURw6 B3LfF75U7a642vtzE/F7enb1Q+czeWnn3ph2w6bEeYCmfbu3cJjd40OQagzyiOvrKuWrlpJa hUWKGKRJZJCv/W65os9qs7KLlSW9Ild6zKVENcFI0WVW0vh3YsVLAAKCGJqWbyB6Mp0J8Vv5 gePpdhZ/Y2LxXxVy+Mq87FuDBjMYrJdebr27lamGphyG99hT1e6Mlk95JLEaZhLM4+wSLxSU zhgw95a5G56t4bBrS2j2sjWJIqq1vLG54zQsZC8n4TqJGgChU1rUNKrK6MVcHBBp1eBtPqHb 2OosRV7nwe08hu6lMFflslt3CTbdwOSzopooJ8oNuJXVNEZA8ZMRm8rRgggg8+xbYfd+9oLP crTfB7e7VHvEdx9SxgiMFvJc6QvjyWsbC3kkUD9N5I3aPihVs9C9ueucJLNrGbmG5kgMfh97 anEdSdCyMDIqknIVgG4Go6GH3NHQT697917r3v3Xuve/de697917r3v3Xuve/de697917r3v 3Xuve/de697917r3v3Xuve/de697917r3v3Xuve/de697917r3v3Xuve/de697917r3v3Xuv e/de697917r3v3Xuve/de6//0N/j37r3Xvfuvde9+691737r3Xvfuvde9+691737r3Xvfuvd e9+691737r3Xvfuvde9+691737r3Xvfuvde9+691737r3Xvfuvde9+691737r3Xvfuvde9+6 91737r3Xvfuvde9+691737r3Xvfuvde9+691737r3Xvfuvde9+691737r3Xvfuvdf//ZAAAA AAA= --------------5600935A196F9FFC789BF743-- --------------1578AA2B782F8AC1EEB779EB--