Aller directement au contenu

NodeBB
  1. Accueil

Monde

Sujets en dehors de ce forum. Les vues et opinions exprimées ici ne reflètent pas nécessairement celles de ce forum et de ses membres.

Un monde de contenus à portée de main…

Voyez cela comme votre fil de découverte global. Il rassemble des discussions intéressantes issues du web et d’autres communautés, en un seul endroit.

Bien que vous puissiez consulter les tendances du moment, la meilleure façon d’utiliser ce fil est de le personnaliser. En créant un compte, vous pouvez suivre des créateurs et des sujets spécifiques afin de filtrer le bruit et de ne voir que ce qui vous intéresse.

Prêt à vous lancer ? Créez un compte pour suivre d’autres personnes, recevoir des notifications quand on vous répond et sauvegarder vos contenus favoris.

S'inscrire Se connecter
  • fariasF
    fariasF farias
    Announcements

    Ancienne version : 2025.7.0

    # bash -c "$(curl -fsSL https://git.community-scripts.org/community-scripts/ProxmoxVE/raw/branch/main/ct/vaultwarden.sh)"

    Core :

    9d6b34a8-179e-4c0c-b3c7-8afb97b4860c-image.png

    e355ce38-df1e-4066-a3f0-16a8e5514247-image.png

    Web :

    08944e98-0bbd-4663-9191-a4e3de16979d-image.png

    4 0 0 Répondre
    fariasF farias
    Mise à jours en cours : [image: image.jpeg]
  • fariasF
    fariasF farias
    Announcements

    Mise à jours de la version de NodeBB via le script :

    ./nodebb stop
./nodebb upgrade
./nodebb start
    35 0 0 Répondre
    fariasF farias
    Mise à jours. # git reset --hard v4.16.0
  • fariasF
    fariasF farias
    Scan Attack
    70 0 0 Répondre
    fariasF farias
    # grep "wp-login.php" /var/log/apache2/access.*.log | sed 's/:/ /g' | awk '{print $2}' | sort -n | uniq -c | sort -n | tail -10 1 90.116.27.237 1 92.180.239.91 1 94.31.70.217 2 174.138.77.239 2 185.19.40.76 2 188.130.129.62 2 45.133.112.95 2 91.132.124.66 13 87.232.150.147 489 150.230.106.64
  • fariasF
    fariasF farias
    Linux

    Visiblement quand je regarde sur https://check.spamhaus.org/troubleshoot/ :

    J’ai deux erreurs :

    dbd97829-c964-4882-89c7-f108a7d31e75-image.png

    Alors que quand je regarde https://mxtoolbox.com/ :

    bd3509d3-8167-4c22-973c-8cb7c2631166-image.png

    Ensuite si je demande un HELO Check :

    Connecting to 80.15.48.50

220 mail.arias-frederic.org ESMTP Postcow [774 ms]
EHLO keeper-us-east-1d.mxtoolbox.com
250-mail.arias-frederic.org
250-PIPELINING
250-SIZE 104857600
250-ETRN
250-STARTTLS
250-ENHANCEDSTATUSCODES
250-8BITMIME
250 DSN [203 ms]
MAIL FROM:<supertool@mxtoolboxsmtpdiag.com>
250 2.1.0 Ok [219 ms]
RCPT TO:<test@mxtoolboxsmtpdiag.com>
454 4.7.1 <test@mxtoolboxsmtpdiag.com>: Relay access denied [297 ms]


- LookupServer [smtp:80.15.48.50] 2743ms

    7b1e8e85-4b46-4823-869b-e7581db8e03d-image.png

    Bref j’ai bien : “mail.arias-frederic.org” et pas “local.balkany” .

    11 0 0 Répondre
    fariasF farias
    Logs : # telnet smtp.arias-frederic.org 25 Trying 80.15.48.50... Connected to smtp.arias-frederic.org. Escape character is '^]'. 220-mail.arias-frederic.org ESMTP Postcow 220 mail.arias-frederic.org ESMTP Postcow HELO testing.com 250 mail.arias-frederic.org EHLO testing.com 250-mail.arias-frederic.org 250-SIZE 104857600 250-ETRN 250-STARTTLS 250-ENHANCEDSTATUSCODES 250 8BITMIME
  • fariasF
    fariasF farias
    Announcements

    J’ai fait un update du PVE : 8.4.11 => 8.4.12.

    Pas de problème.

    36 0 0 Répondre
    fariasF farias
    Update : [image: image.jpeg]
  • Tuxedo17T
    Tuxedo17T Tuxedo17
    IA

    Méthode d’identification de la compression :

    Le prompt “You are a summarization agent…” n’est pas stocké dans la table activity de llama-swap (elle ne garde que des métriques, pas le contenu des requêtes). J’ai donc identifié les requêtes de
    compression par leur signature : un ratio input_tokens/output_tokens > 100 (un checkpoint énorme en entrée, un résultat de synthèse court en sortie).

    Résultats clés :

    Métrique Valeur
    Période 22/07 14h → 06/08 17h (85 heures)
    Temps total 72,40h
    Temps compression 16,61h (22,9%)
    Temps normal 55,79h (77,1%)
    Pic compression 98% à 22/07 16h
    Pic moyen/jour 32% à 28/07

    Les deux graphiques montrent :

    1. Haut : empilement des temps (barres bleues = normal, oranges = compression)
    2. Bas : pourcentage de compression par heure/jour

    Le graphique *-daily.png est plus lisible pour une vue d’ensemble, tandis que *-hourly.png donne le détail fin par heure.

    llama-swap-compression-daily.png
    llama-swap-compression-hourly.png

    4 0 0 Répondre
    Tuxedo17T Tuxedo17
    Update : [image: llama-swap-compression-hourly.png]
  • fariasF
    fariasF farias
    Announcements 
    # cd /www/pixelfed/pixelfed
# systemctl stop pixelfed
# rm composer.lock
...
# composer install --no-ansi --no-dev --no-interaction --no-progress --no-scripts --optimize-autoloader
...
Generating optimized autoload files
Class App\Rules\WebFinger located in ./app/Rules/Webfinger.php does not comply with psr-4 autoloading standard (rule: App\ => ./app). Skipping.
83 packages you are using are looking for funding.
Use the `composer fund` command to find out more!

# composer dump-autoload --optimize
...
#  php artisan config:cache

   INFO  Configuration cached successfully.  

# php artisan route:cache

   INFO  Routes cached successfully.  

# php artisan migrate --force

   INFO  Running migrations.  

  2025_08_30_044247_fix_stories_table_set_view_counts_default ......................................................................... 77.42ms DONE
# chown -R pixelfed:pixelfed /www/pixelfed/pixelfed/
# systemctl start pixelfed

    Bref j’ai eu une erreur, si je regarde : ./app/Rules/Webfinger.php

    <?php

namespace App\Rules;

use Illuminate\Contracts\Validation\Rule;

    Si je regarde les namespace dans les PHP

    # grep "^namespace" ./app/*/*.* | awk '{print $2}' | sort -n | uniq -c
      1 App\Auth;
      1 App\Console;
      1 App\Events;
      1 App\Exceptions;
      1 App\Http;
      2 App\Listeners;
     19 App\Mail;
     53 App\Models;
     11 App\Observers;
      1 App\Policies;
      8 App\Providers;
      6 App\Rules;
     69 App\Services;

    Dans composer.json j’ai :

            "psr-4": {
            "App\\": "app/"
        },

    je comprends pas le problème mais la mise à jours semble fonctionner.

    5 0 0 Répondre
    fariasF farias
    Update de composer : # composer self-update Upgrading to version 2.10.3 (stable channel). Use composer self-update --rollback to return to version 2.9.5 la mise à jours : # cd /www/pixelfed/pixelfed # systemctl stop pixelfed # composer install --no-ansi --no-dev --no-interaction --no-progress --no-scripts --optimize-autoloader Composer plugins have been disabled for safety in this non-interactive session. Set COMPOSER_ALLOW_SUPERUSER=1 if you want to allow plugins to run as root/super user. Do not run Composer as root/super user! See https://getcomposer.org/root for details Installing dependencies from lock file Verifying lock file contents can be installed on current platform. Nothing to install, update or remove Generating optimized autoload files Class App\Rules\WebFinger located in ./app/Rules/Webfinger.php does not comply with psr-4 autoloading standard (rule: App\ => ./app). Skipping. 85 packages you are using are looking for funding. Use the `composer fund` command to find out more! root@insta:/www/pixelfed/pixelfed# composer dump-autoload --optimize Do not run Composer as root/super user! See https://getcomposer.org/root for details Continue as root/super user [yes]? yes Generating optimized autoload files Class App\Rules\WebFinger located in ./app/Rules/Webfinger.php does not comply with psr-4 autoloading standard (rule: App\ => ./app). Skipping. > Illuminate\Foundation\ComposerScripts::postAutoloadDump > @php artisan package:discover --ansi INFO Discovering packages. buzz/laravel-h-captcha ...................................................................................................................... DONE intervention/image-laravel .................................................................................................................. DONE jenssegers/agent ............................................................................................................................ DONE laravel-notification-channels/expo .......................................................................................................... DONE laravel-notification-channels/webpush ....................................................................................................... DONE laravel/horizon ............................................................................................................................. DONE laravel/pulse ............................................................................................................................... DONE laravel/tinker .............................................................................................................................. DONE laravel/ui .................................................................................................................................. DONE livewire/livewire ........................................................................................................................... DONE nesbot/carbon ............................................................................................................................... DONE nunomaduro/termwind ......................................................................................................................... DONE pbmedia/laravel-ffmpeg ...................................................................................................................... DONE pixelfed/laravel-snowflake .................................................................................................................. DONE spatie/laravel-backup ....................................................................................................................... DONE spatie/laravel-image-optimizer .............................................................................................................. DONE spatie/laravel-signal-aware-command ......................................................................................................... DONE stevebauman/purify .......................................................................................................................... DONE Generated optimized autoload files containing 10045 classes # php artisan config:cache INFO Configuration cached successfully. # php artisan route:cache INFO Routes cached successfully. # php artisan migrate --force INFO Nothing to migrate. # chown -R pixelfed:pixelfed /www/pixelfed/pixelfed/ # systemctl start pixelfed
  • Tuxedo17T
    Tuxedo17T Tuxedo17
    Linux

    Avant :

    $ go version
go version go1.22.2 linux/amd64

    Après le but est l’installation de 1.27.1 à la place de 1.22.2 :

    # go version
go version go1.27.1 linux/amd64

    Le but est l’installation de https://github.com/josephburnett/jd

    # wget -q https://go.dev/dl/go1.27.1.linux-amd64.tar.gz
    0 0 0 Répondre
  • Tuxedo17T
    Tuxedo17T Tuxedo17
    IA

    Dans les logs :

    31.10.328.437 I slot print_timing: id  1 | task 26420 |    graphs reused =      26082
31.10.331.740 I slot      release: id  1 | task 26420 | stop processing: n_tokens = 55637, truncated = 0
32.10.441.577 I srv    operator(): operator(): cleaning up before exit...

    Dans le syslog, j’ai :

    [64949.879619] ------------[ cut here ]------------
[64949.879627] i915 0000:00:02.0: [drm] drm_WARN_ON_ONCE(t_vblank < vblank->time)
[64949.879631] WARNING: drivers/gpu/drm/drm_vblank.c:1572 at drm_vblank_restore+0x205/0x280, CPU#7: eDP-1/1662
[64949.879646] Modules linked in: tcp_diag inet_diag tls ccm rfcomm snd_seq_dummy snd_hrtimer evdi(OE) nvidia_uvm(OE) qrtr nvidia(OE) cmac algif_hash algif_skcipher af_alg bnep typec_displayport typec_thunderbolt snd_sof_pci_intel_tgl snd_sof_pci_intel_cnl snd_sof_intel_hda_generic soundwire_intel snd_sof_intel_hda_sdw_bpt snd_sof_intel_hda_common xe snd_soc_hdac_hda snd_sof_intel_hda_mlink snd_sof_intel_hda soundwire_cadence snd_sof_pci snd_sof_xtensa_dsp snd_sof snd_sof_utils snd_soc_acpi_intel_match snd_soc_acpi_intel_sdca_quirks soundwire_generic_allocation snd_soc_sdw_utils snd_soc_acpi soundwire_bus drm_gpusvm_helper snd_soc_sdca gpu_sched crc8 drm_gpuvm drm_exec drm_suballoc_helper snd_soc_avs snd_soc_hda_codec snd_hda_ext_core snd_hda_codec snd_hda_core snd_intel_dspcfg snd_intel_sdw_acpi snd_hwdep intel_rapl_msr intel_rapl_common snd_soc_core intel_uncore_frequency intel_uncore_frequency_common intel_tcc_cooling snd_compress ac97_bus x86_pkg_temp_thermal snd_pcm_dmaengine intel_powerclamp snd_pcm coretemp
[64949.879750]  snd_seq_midi iwlmvm kvm_intel uvcvideo snd_seq_midi_event snd_rawmidi mac80211 videobuf2_vmalloc snd_seq uvc videobuf2_memops snd_seq_device sunrpc videobuf2_v4l2 libarc4 snd_timer cmdlinepart tuxedo_nb02_nvidia_power_ctrl(OE) videobuf2_common btusb kvm spi_nor uniwill_wmi(OE) ee1004 i915 btmtk binfmt_misc tuxedo_io(OE) clevo_wmi(OE) videodev mei_hdcp snd mei_pxp iwlwifi btrtl irqbypass mtd tuxedo_keyboard(OE) asus_wmi i2c_i801 btbcm rapl tuxedo_compatibility_check(OE) btintel i2c_smbus drm_buddy mei_me spi_intel_pci nls_iso8859_1 intel_cstate platform_profile wmi_bmof led_class_multicolor mc soundcore i2c_mux cfg80211 bluetooth spi_intel drm_display_helper mei nvidiafb vgastate cec fb_ddc rc_core intel_pmc_core i2c_algo_bit pmt_telemetry pmt_discovery pmt_class intel_pmc_ssram_telemetry acpi_pad intel_hid intel_vsec acpi_tad sparse_keymap joydev input_leds mac_hid serio_raw sch_fq_codel vmwgfx drm_ttm_helper ttm msr parport_pc ppdev lp parport dm_crypt nvme_fabrics efi_pstore nfnetlink dmi_sysfs
[64949.879885]  ip_tables x_tables autofs4 btrfs libblake2b raid10 raid456 async_raid6_recov async_memcpy async_pq async_xor async_tx xor raid6_pq raid1 raid0 linear usbhid cdc_ether usbnet r8152 mii ucsi_acpi typec_ucsi typec hid_multitouch r8125(OE) hid_generic nvme uas usb_storage nvme_core ghash_clmulni_intel r8169 thunderbolt i2c_hid_acpi intel_lpss_pci ahci nvme_keyring i2c_hid intel_lpss realtek video nvme_auth libahci idma64 hid hkdf wmi pinctrl_tigerlake aesni_intel
[64949.879970] CPU: 7 UID: 1000 PID: 1662 Comm: eDP-1 Tainted: G           OE       7.0.0-110030-tuxedo #30~24.04.1tux1 PREEMPT(lazy) 
[64949.879975] Tainted: [O]=OOT_MODULE, [E]=UNSIGNED_MODULE
[64949.879976] Hardware name: TUXEDO TUXEDO Polaris Intel Gen3 (TGL)/GMxTGxx, BIOS N.1.07A03 09/24/2021
[64949.879979] RIP: 0010:drm_vblank_restore+0x20f/0x280
[64949.879983] Code: 48 8b 7b 08 4c 8b 67 50 4d 85 e4 74 73 4c 89 4d b8 e8 95 22 ec ff 48 89 c6 48 8d 3d fb 59 e0 01 48 c7 c1 a0 61 d7 b0 4c 89 e2 <67> 48 0f b9 3a 45 31 c0 4c 8b 4d b8 45 89 f4 e9 e3 fe ff ff 85 c0
[64949.879985] RSP: 0018:ffffd2c10448b718 EFLAGS: 00010046
[64949.879988] RAX: ffffffffc1a900a0 RBX: ffff8a0cce854000 RCX: ffffffffb0d761a0
[64949.879990] RDX: ffff8a0cc3671c80 RSI: ffffffffc1a900a0 RDI: ffffffffb1cfcf50
[64949.879991] RBP: ffffd2c10448b760 R08: 0000000000000000 R09: fffffffffffff60f
[64949.879993] R10: 0000000000000000 R11: 0000000000000000 R12: ffff8a0cc3671c80
[64949.879994] R13: 0000000000000003 R14: 00000000000b790b R15: ffff8a0cc3958828
[64949.879996] FS:  0000799d325fd6c0(0000) GS:ffff8a1c7d563000(0000) knlGS:0000000000000000
[64949.879998] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[64949.880000] CR2: 00000000284b4398 CR3: 0000000131fcf001 CR4: 0000000000f72ef0
[64949.880002] PKRU: 55555554
[64949.880003] Call Trace:
[64949.880005]  <TASK>
[64949.880011]  drm_crtc_vblank_restore+0x6c/0x150
[64949.880017]  bdw_enable_vblank+0xd8/0x100 [i915]
[64949.880285]  drm_vblank_enable+0x104/0x1b0
[64949.880290]  drm_vblank_get+0xb7/0x100
[64949.880294]  drm_crtc_vblank_get+0x17/0x30
[64949.880298]  intel_display_rps_boost_after_vblank+0x58/0x130 [i915]
[64949.880492]  intel_prepare_plane_fb+0xfa/0x160 [i915]
[64949.880695]  drm_atomic_helper_prepare_planes.part.0+0x32/0x1e0
[64949.880701]  drm_atomic_helper_prepare_planes+0x5d/0x70
[64949.880705]  intel_atomic_commit+0x145/0x310 [i915]
[64949.880898]  drm_atomic_nonblocking_commit+0x5c/0x70
[64949.880902]  drm_mode_atomic_ioctl+0x761/0x8a0
[64949.880907]  ? __pfx_drm_mode_atomic_ioctl+0x10/0x10
[64949.880911]  drm_ioctl_kernel+0xb7/0x110
[64949.880915]  drm_ioctl+0x2ec/0x5b0
[64949.880919]  ? __pfx_drm_mode_atomic_ioctl+0x10/0x10
[64949.880924]  __x64_sys_ioctl+0xa5/0x100
[64949.880929]  x64_sys_call+0x1250/0x26e0
[64949.880933]  do_syscall_64+0xd9/0x5b0
[64949.880938]  ? psi_task_switch+0x130/0x3a0
[64949.880943]  ? raw_spin_rq_unlock+0x14/0x50
[64949.880947]  ? finish_task_switch.isra.0+0x95/0x2d0
[64949.880951]  ? __schedule+0x2d8/0x660
[64949.880954]  ? schedule+0x3d/0x90
[64949.880956]  ? do_nanosleep+0x68/0x180
[64949.880960]  ? hrtimer_setup_sleeper_on_stack+0x2e/0xb0
[64949.880965]  ? hrtimer_nanosleep+0x8b/0x110
[64949.880967]  ? __pfx_hrtimer_wakeup+0x10/0x10
[64949.880971]  ? common_nsleep+0x43/0x60
[64949.880975]  ? __x64_sys_clock_nanosleep+0xf2/0x180
[64949.880977]  ? note_gp_changes+0x8f/0xa0
[64949.880981]  ? exit_to_user_mode_loop+0x39d/0x540
[64949.880985]  ? arch_exit_to_user_mode_prepare.isra.0+0xc3/0xe0
[64949.880988]  ? do_syscall_64+0x116/0x5b0
[64949.880992]  ? handle_softirqs+0xe5/0x360
[64949.880996]  ? arch_exit_to_user_mode_prepare.isra.0+0xd/0x100
[64949.880999]  ? irqentry_exit+0x9d/0x640
[64949.881001]  ? clear_bhb_loop+0x50/0xa0
[64949.881006]  entry_SYSCALL_64_after_hwframe+0x76/0x7e
[64949.881008] RIP: 0033:0x799d4bd24f1d
[64949.881011] Code: 04 25 28 00 00 00 48 89 45 c8 31 c0 48 8d 45 10 c7 45 b0 10 00 00 00 48 89 45 b8 48 8d 45 d0 48 89 45 c0 b8 10 00 00 00 0f 05 <89> c2 3d 00 f0 ff ff 77 1a 48 8b 45 c8 64 48 2b 04 25 28 00 00 00
[64949.881013] RSP: 002b:0000799d325fc200 EFLAGS: 00000246 ORIG_RAX: 0000000000000010
[64949.881016] RAX: ffffffffffffffda RBX: 0000799d280054e0 RCX: 0000799d4bd24f1d
[64949.881017] RDX: 0000799d325fc2f0 RSI: 00000000c03864bc RDI: 0000000000000012
[64949.881019] RBP: 0000799d325fc250 R08: 0000799d28002bb0 R09: 0000799d28002b20
[64949.881020] R10: 0000799d2800cb30 R11: 0000000000000246 R12: 0000799d325fc2f0
[64949.881021] R13: 00000000c03864bc R14: 0000000000000012 R15: 0000000000000008
[64949.881024]  </TASK>
[64949.881025] ---[ end trace 0000000000000000 ]---

    Je ne sais pas si cela a un rapport.

    0 0 0 Répondre
  • Tuxedo17T
    Tuxedo17T Tuxedo17
    IA

    Chargement : https://www.nvidia.com/en-us/drivers/details/277983/

    a2cd2f95-3699-4b60-8847-23634898fa31-image.jpeg

    2 0 0 Répondre
    Tuxedo17T Tuxedo17
    Pas de crash : # nvidia-smi Tue Sep 1 17:35:06 2026 +-----------------------------------------------------------------------------------------+ | NVIDIA-SMI 610.57.04 KMD Version: 610.57.04 CUDA UMD Version: 13.3 | +-----------------------------------------+------------------------+----------------------+ | GPU Name Persistence-M | Bus-Id Disp.A | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap | Memory-Usage | GPU-Util Compute M. | | | | MIG M. | |=========================================+========================+======================| | 0 NVIDIA GeForce RTX 3060 ... Off | 00000000:01:00.0 Off | N/A | | N/A 63C P0 36W / 115W | 2096MiB / 6144MiB | 12% Default | | | | N/A | +-----------------------------------------+------------------------+----------------------+ | 1 NVIDIA GeForce RTX 5060 Ti Off | 00000000:05:00.0 Off | N/A | | 30% 51C P1 49W / 180W | 7894MiB / 16311MiB | 32% Default | | | | N/A | +-----------------------------------------+------------------------+----------------------+ +-----------------------------------------------------------------------------------------+ | Processes: | | GPU GI CI PID Type Process name GPU Memory | | ID ID Usage | |=========================================================================================| | 0 N/A N/A 3587 C /usr/local/bin/llama-server 2088MiB | | 1 N/A N/A 3587 C /usr/local/bin/llama-server 7886MiB | +-----------------------------------------------------------------------------------------+
  • Tuxedo17T
    Tuxedo17T Tuxedo17
    IA

    Le dernier logs avant le crash :

    sept. 01 14:24:37.636400 tuxedo-17 systemd[1]: systemd 255.4-1ubuntu8.17 running in system mode (+PAM +AUDIT +SELINUX +APPARMOR +IMA +SMACK +SECCOMP +GCRYPT -GNUTLS +OPENSSL +ACL +BLKID +CURL +ELFUTILS +FIDO2 +IDN2 -IDN +IPTC +KMOD +LI>
sept. 01 14:24:37.636491 tuxedo-17 systemd[1]: Detected architecture x86-64.
sept. 01 14:24:37.685390 tuxedo-17 systemd[1]: getty.target: Wants dependency dropin /etc/systemd/system/getty.target.wants/getty@tty1.service target /lib/systemd/system/autovt@.service has different name
sept. 01 14:24:37.694384 tuxedo-17 systemd[1]: /etc/systemd/system/llama-swap.service:12: Unknown key name 'Environnent' in section 'Service', ignoring.
sept. 01 14:24:37.759392 tuxedo-17 systemd[1]: /etc/systemd/system/llama-server.service:12: Unknown key name 'Environnent' in section 'Service', ignoring.
sept. 01 14:24:37.873862 tuxedo-17 systemd-journald[421]: Received SIGTERM from PID 1 (systemd).
sept. 01 14:24:37.873893 tuxedo-17 systemd[1]: Stopping systemd-journald.service - Journal Service...
sept. 01 14:24:37.873908 tuxedo-17 systemd[1]: systemd-journald.service: Deactivated successfully.
sept. 01 14:24:37.873916 tuxedo-17 systemd[1]: Stopped systemd-journald.service - Journal Service.
sept. 01 14:24:37.873924 tuxedo-17 systemd[1]: systemd-journald.service: Consumed 17min 2.264s CPU time, 28.9M memory peak, 0B memory swap peak.
sept. 01 14:24:37.873932 tuxedo-17 systemd[1]: Starting systemd-journald.service - Journal Service...
sept. 01 14:24:37.873939 tuxedo-17 systemd-journald[984342]: Collecting audit messages is disabled.
sept. 01 14:24:37.874382 tuxedo-17 systemd[1]: Started systemd-journald.service - Journal Service.
sept. 01 14:26:59.886416 tuxedo-17 kernel: /proc/cgroups lists only v1 controllers, use cgroup.controllers of root cgroup for v2 info
sept. 01 14:27:05.451421 tuxedo-17 kernel: NVRM: API mismatch: the client 'nvidia-smi' (pid 1008648)
                                           NVRM: has the version 580.178.04, but this kernel module has
                                           NVRM: the version 610.43.03.  Please make sure that this
                                           NVRM: kernel module and all NVIDIA driver components
                                           NVRM: have the same version.
sept. 01 14:27:05.460389 tuxedo-17 kernel: NVRM: API mismatch: the client 'nvidia-smi' (pid 1008651)
                                           NVRM: has the version 580.178.04, but this kernel module has
                                           NVRM: the version 610.43.03.  Please make sure that this
                                           NVRM: kernel module and all NVIDIA driver components
                                           NVRM: have the same version.
sept. 01 14:27:16.792389 tuxedo-17 kernel: wlo1: deauthenticating from c2:55:88:4e:1b:b2 by local choice (Reason: 3=DEAUTH_LEAVING)
sept. 01 14:27:20.460396 tuxedo-17 kernel: wlo1: authenticate with c2:55:b8:4e:1b:b2 (local address=14:85:7f:b5:79:3f)
sept. 01 14:27:20.461397 tuxedo-17 kernel: wlo1: send auth to c2:55:b8:4e:1b:b2 (try 1/3)
sept. 01 14:27:20.491409 tuxedo-17 kernel: wlo1: authenticated
sept. 01 14:27:20.491434 tuxedo-17 kernel: wlo1: associate with c2:55:b8:4e:1b:b2 (try 1/3)
sept. 01 14:27:20.499406 tuxedo-17 kernel: wlo1: RX AssocResp from c2:55:b8:4e:1b:b2 (capab=0x1511 status=0 aid=2)
sept. 01 14:27:20.507381 tuxedo-17 kernel: wlo1: associated
sept. 01 14:27:20.514388 tuxedo-17 kernel: iwlwifi 0000:00:14.3: Unhandled alg: 0x707
sept. 01 14:33:58.182481 tuxedo-17 kernel: workqueue: inode_switch_wbs_work_fn hogged CPU for >10000us 4 times, consider switching to WQ_UNBOUND
sept. 01 14:33:58.213408 tuxedo-17 kernel: workqueue: inode_switch_wbs_work_fn hogged CPU for >10000us 5 times, consider switching to WQ_UNBOUND
sept. 01 14:33:58.231409 tuxedo-17 kernel: workqueue: inode_switch_wbs_work_fn hogged CPU for >10000us 7 times, consider switching to WQ_UNBOUND

    Après le reboot :

    # nvidia-smi 
Tue Sep  1 14:50:52 2026       
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 610.43.03              KMD Version: 610.43.03     CUDA UMD Version: 13.3     |
+-----------------------------------------+------------------------+----------------------+
| GPU  Name                 Persistence-M | Bus-Id          Disp.A | Volatile Uncorr. ECC |
| Fan  Temp   Perf          Pwr:Usage/Cap |           Memory-Usage | GPU-Util  Compute M. |
|                                         |                        |               MIG M. |
|=========================================+========================+======================|
|   0  NVIDIA GeForce RTX 3060 ...    Off |   00000000:01:00.0 Off |                  N/A |
| N/A   47C    P8             11W /  115W |       0MiB /   6144MiB |      0%      Default |
|                                         |                        |                  N/A |
+-----------------------------------------+------------------------+----------------------+
|   1  NVIDIA GeForce RTX 5060 Ti     Off |   00000000:05:00.0 Off |                  N/A |
|  0%   43C    P8              5W /  180W |       0MiB /  16311MiB |      0%      Default |
|                                         |                        |                  N/A |
+-----------------------------------------+------------------------+----------------------+

+-----------------------------------------------------------------------------------------+
| Processes:                                                                              |
|  GPU   GI   CI              PID   Type   Process name                        GPU Memory |
|        ID   ID                                                               Usage      |
|=========================================================================================|
|  No running processes found                                                             |
+-----------------------------------------------------------------------------------------+
    4 0 0 Répondre
    Tuxedo17T Tuxedo17
    C’est mieux : [image: 7f991335-22e6-424f-9fe4-4495751d1e66-image.jpeg]
  • fariasF
    fariasF farias
    Announcements 
    # python3.11 -m venv garmin-to-fittrackee
# source garmin-to-fittrackee/bin/activate
# cd garmin-to-fittrackee/
# git clone https://git.dryusdan.fr/Dryusdan/garmin-to-fittrackee.git
# cd garmin-to-fittrackee
# pip install poetry
# poetry install

Current Python version (3.11.2) is not allowed by the project (^3.12).
Please change python executable via the "env use" command.
# cat /etc/os-release 
PRETTY_NAME="Debian GNU/Linux 12 (bookworm)"
NAME="Debian GNU/Linux"
VERSION_ID="12"
VERSION="12 (bookworm)"
VERSION_CODENAME=bookworm
ID=debian
HOME_URL="https://www.debian.org/"
SUPPORT_URL="https://www.debian.org/support"
BUG_REPORT_URL="https://bugs.debian.org/"
# deactivate
    4 0 0 Répondre
    fariasF farias
    C’est bon cela fonctionne à nouveau … ERROR [sports/get_fittrackee_sport_by_garmin_id] Not Fittrackee match with garmin sport id 240 sports.py:134 ERROR [sports/get_fittrackee_sport_by_garmin_id] Set to cycling road sports.py:137 INFO [main/_fetch_garmin_activity_file] Activity data downloaded to file /tmp/24145714013.zip main.py:241 [08/31/26 13:35:04] INFO [fittrackee/upload_workout] Activity added on Fittrackee with id KdjFEeZJPpk24idGwDRfSS fittrackee.py:288 INFO [main/sync] Fetching activities on Garminfrom Aug 30, 2026to Aug 31, 2026 main.py:163 [08/31/26 13:35:05] INFO [main/_fetch_garmin_activity_file] Activity data downloaded to file /tmp/24181394659.zip main.py:241 [08/31/26 13:35:20] INFO [fittrackee/upload_workout] Activity added on Fittrackee with id G8EdcT5tnX29hNE7ucuwte fittrackee.py:288 Le surf est vue comme du vélo … misère.
  • fariasF
    fariasF farias
    Announcements

    Passage v0.11.2 => v0.12.0.

    Pour avoir les flux RSS.

    13 0 0 Répondre
    fariasF farias
    Update done : $ ftcli db upgrade INFO [alembic.runtime.migration] Context impl PostgresqlImpl. INFO [alembic.runtime.migration] Will assume transactional DDL. INFO [alembic.runtime.migration] Running upgrade 3d0c336b7a9e -> 2e3a59ebbc59, update blacklisted_tokens table
  • Tuxedo17T
    Tuxedo17T Tuxedo17
    IA

    Mon architecture :

     nvidia-smi 
Fri Aug  7 15:51:53 2026       
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 610.43.03              KMD Version: 610.43.03     CUDA UMD Version: 13.3     |
+-----------------------------------------+------------------------+----------------------+
| GPU  Name                 Persistence-M | Bus-Id          Disp.A | Volatile Uncorr. ECC |
| Fan  Temp   Perf          Pwr:Usage/Cap |           Memory-Usage | GPU-Util  Compute M. |
|                                         |                        |               MIG M. |
|=========================================+========================+======================|
|   0  NVIDIA GeForce RTX 3060 ...    Off |   00000000:01:00.0 Off |                  N/A |
| N/A   64C    P0             33W /  115W |    1918MiB /   6144MiB |      8%      Default |
|                                         |                        |                  N/A |
+-----------------------------------------+------------------------+----------------------+
|   1  NVIDIA GeForce RTX 5060 Ti     Off |   00000000:05:00.0 Off |                  N/A |
| 30%   51C    P1             41W /  180W |    7550MiB /  16311MiB |     25%      Default |
|                                         |                        |                  N/A |
+-----------------------------------------+------------------------+----------------------+

+-----------------------------------------------------------------------------------------+
| Processes:                                                                              |
|  GPU   GI   CI              PID   Type   Process name                        GPU Memory |
|        ID   ID                                                               Usage      |
|=========================================================================================|
|    0   N/A  N/A         1863585      C   /usr/local/bin/llama-server            1910MiB |
|    1   N/A  N/A         1863585      C   /usr/local/bin/llama-server            7542MiB |
+-----------------------------------------------------------------------------------------+

    Ma configuration :

    /usr/local/bin/llama-server -m /models/Qwen3.6-35B-A3B-UD-Q4_K_M.gguf --ctx-size ${default_ctx} --temp ${temp} --top-p 0.95 --top-k 20 --min-p 0.00  --reasoning-preserve  --host 0.0.0.0 --fit off -ngl 99  --port 8999 --n-cpu-moe 35 -fa on --flash-attn on

    Résultat :

    llama-swap-prompt-speed-daily.png

    2 0 1 Répondre
    Tuxedo17T Tuxedo17
    Je viens de refaire un graphique : [image: llama-swap-prompt-speed-daily.png]
  • Tuxedo17T
    Tuxedo17T Tuxedo17
    DSI

    https://infomaxparis.com/fr/pc-professionnel/32530-407614-pc-creator-ultimate-nvidia-studio.html : 5000 Euros.

    6 0 0 Répondre
    Tuxedo17T Tuxedo17
    Voici le panier : https://ldlc.com/s/3PYTTLK
  • Tuxedo17T
    Tuxedo17T Tuxedo17
    IA

    Site : https://www.demeter-sante.fr/calculateur-llm
    Mes valeurs :
    2a3b6004-ac63-418d-b6d0-c8146908cd5d-image.jpeg

    Cached : 312 405 987 tokens
    Processed : 19 792 357 tokens
    Generated : 3 288 290 tokens

    IN : Cached+Processed = 332198344 ?
    OUT : Generated : 3288290 ?

    90bc3bcb-6042-4232-a229-83f94d8fc128-image.jpeg

    0 0 1 Répondre
  • fariasF
    fariasF farias
    Dawarich

    Voici les ressources :

    • 260 Go d’espace disque ( utilisé à 90 % )

    5c8ab908-b80e-4f35-ba50-e4a23fd56d54-image.jpeg

    ceaf2c4f-7fb6-4477-9f0d-0f2778e42a61-image.jpeg

    e2813bc9-df25-4bb2-bd78-abb7b86d5cf8-image.jpeg

    c4d00280-907e-4638-9798-c24db95d09a1-image.jpeg

    ca5a037c-e2d1-4cbb-a8a6-37fe87cfc609-image.jpeg

    85039fec-8d40-42f2-8a99-c3513a49e0d7-image.jpeg

    1cdec5b3-4602-4b3c-a032-b29d11677e43-image.jpeg

    0 0 0 Répondre
  • Tuxedo17T
    Tuxedo17T Tuxedo17
    IA

    Installation de vllm : https://oneuptime.com/blog/post/2026-03-02-how-to-install-and-configure-vllm-on-ubuntu/view

    Step 1: Verify GPU Setup

    # nvidia-smi
Fri Aug  7 11:33:15 2026       
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 610.43.03              KMD Version: 610.43.03     CUDA UMD Version: 13.3     |
+-----------------------------------------+------------------------+----------------------+
| GPU  Name                 Persistence-M | Bus-Id          Disp.A | Volatile Uncorr. ECC |
| Fan  Temp   Perf          Pwr:Usage/Cap |           Memory-Usage | GPU-Util  Compute M. |
|                                         |                        |               MIG M. |
|=========================================+========================+======================|
|   0  NVIDIA GeForce RTX 3060 ...    Off |   00000000:01:00.0 Off |                  N/A |
| N/A   46C    P8             15W /  115W |       0MiB /   6144MiB |      0%      Default |
|                                         |                        |                  N/A |
+-----------------------------------------+------------------------+----------------------+
|   1  NVIDIA GeForce RTX 5060 Ti     Off |   00000000:05:00.0 Off |                  N/A |
|  0%   45C    P8              5W /  180W |       0MiB /  16311MiB |      0%      Default |
|                                         |                        |                  N/A |
+-----------------------------------------+------------------------+----------------------+

+-----------------------------------------------------------------------------------------+
| Processes:                                                                              |
|  GPU   GI   CI              PID   Type   Process name                        GPU Memory |
|        ID   ID                                                               Usage      |
|=========================================================================================|
|  No running processes found                                                             |
+-----------------------------------------------------------------------------------------+

# nvcc --version || nvidia-smi | grep "CUDA Version"
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2023 NVIDIA Corporation
Built on Fri_Jan__6_16:45:21_PST_2023
Cuda compilation tools, release 12.0, V12.0.140
Build cuda_12.0.r12.0/compiler.32267302_0

    Step 2: Create Python Environment

    # Install Python, venv, curl, and uv
sudo apt-get update && sudo apt-get install -y python3 python3-pip python3-venv curl
curl -LsSf https://astral.sh/uv/install.sh | sh
source "$HOME/.local/bin/env"

# Create a dedicated venv
uv venv --python 3.12 --seed --managed-python ~/vllm-env
source ~/vllm-env/bin/activate

# Upgrade pip
pip install --upgrade pip

    Step 3: Install vLLM

    # Install vLLM with a PyTorch backend selected for your CUDA driver
uv pip install vllm --torch-backend=auto

# Verify installation
python3 -c "import vllm; print(vllm.__version__)"

    Step 4: Download a Model

    # Install huggingface_hub for model downloading
pip install huggingface_hub

# Pre-download a model (optional - vLLM downloads on first use)
python3 -c "
from huggingface_hub import snapshot_download
snapshot_download(
    repo_id='meta-llama/Meta-Llama-3-8B-Instruct',  # 8B model, needs 16GB VRAM
    local_dir='/models/llama-3.2-1b-instruct'
)
"

    Je vais devoir ajouter dans mon config.xml :

    host: "127.0.0.1"
port: 8080
idle_ttl: 300 

models:
  # Modèle : Nvidia Qwen 3.6 35B NVFP4 poussé à 180k de contexte
  - name: "nvidia/Qwen3.6-35B-A3B-NVFP4"
    command: >
      bash -c "source ~/vllm-env/bin/activate && vllm serve nvidia/Qwen3.6-35B-A3B-NVFP4 --port 8002 --gpu-memory-utilization 0.95 --max-model-len 180000 --max-num-seqs 1 --kv-cache-dtype fp8 --enable-chunked-prefill --max-num-batched-tokens 4096 --gpu-memory-buffer-kv-cache-space 4 --attention-backend flashinfer --reasoning-parser qwen3"
    ready_url: "http://127.0.0"
    upstream_url: "http://127.0.0"

    A suivre.

    3 0 0 Répondre
    Tuxedo17T Tuxedo17
    Misère … # /root/vllm-env/bin/vllm serve nvidia/Qwen3.6-35B-A3B-NVFP4 --port 8002 --gpu-memory-utilization 0.95 --max-model-len 180000 --max-num-seqs 1 --kv-cache-dtype fp8 --enable-chunked-prefill --max-num-batched-tokens 4096 --attention-backend flashinfer --reasoning-parser qwen3 WARNING 08-07 13:29:38 [cuda.py:959] Detected different devices in the system: NVIDIA GeForce RTX 3060 Laptop GPU, NVIDIA GeForce RTX 5060 Ti. Please make sure to set `CUDA_DEVICE_ORDER=PCI_BUS_ID` to avoid unexpected behavior. (APIServer pid=1833126) INFO 08-07 13:29:49 [api_utils.py:345] (APIServer pid=1833126) INFO 08-07 13:29:49 [api_utils.py:345] █ █ █▄ ▄█ (APIServer pid=1833126) INFO 08-07 13:29:49 [api_utils.py:345] ▄▄ ▄█ █ █ █ ▀▄▀ █ version 0.26.0 (APIServer pid=1833126) INFO 08-07 13:29:49 [api_utils.py:345] █▄█▀ █ █ █ █ model nvidia/Qwen3.6-35B-A3B-NVFP4 (APIServer pid=1833126) INFO 08-07 13:29:49 [api_utils.py:345] ▀▀ ▀▀▀▀▀ ▀▀▀▀▀ ▀ ▀ (APIServer pid=1833126) INFO 08-07 13:29:49 [api_utils.py:345] (APIServer pid=1833126) INFO 08-07 13:29:49 [api_utils.py:273] non-default args: {'model_tag': 'nvidia/Qwen3.6-35B-A3B-NVFP4', 'port': 8002, 'model': 'nvidia/Qwen3.6-35B-A3B-NVFP4', 'max_model_len': 180000, 'attention_backend': 'flashinfer', 'reasoning_parser': 'qwen3', 'gpu_memory_utilization': 0.95, 'kv_cache_dtype': 'fp8', 'max_num_batched_tokens': 4096, 'max_num_seqs': 1, 'enable_chunked_prefill': True} (APIServer pid=1833126) Warning: You are sending unauthenticated requests to the HF Hub. Please set a HF_TOKEN to enable higher rate limits and faster downloads. config.json: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 58.1k/58.1k [00:00<00:00, 93.7MB/s] preprocessor_config.json: 100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 390/390 [00:00<00:00, 2.29MB/s] (APIServer pid=1833126) INFO 08-07 13:30:04 [model.py:623] Resolved architecture: Qwen3_5MoeForConditionalGeneration (APIServer pid=1833126) INFO 08-07 13:30:04 [model.py:1788] Using max model len 180000 (APIServer pid=1833126) INFO 08-07 13:30:05 [cache.py:285] Using fp8 data type to store kv cache. It reduces the GPU memory footprint and boosts the performance. Meanwhile, it may cause accuracy drop without a proper scaling factor (APIServer pid=1833126) INFO 08-07 13:30:05 [scheduler.py:252] Chunked prefill is enabled with max_num_batched_tokens=4096. (APIServer pid=1833126) WARNING 08-07 13:30:05 [modelopt.py:385] Detected ModelOpt fp8 checkpoint (quant_algo=FP8). Please note that the format is experimental and could change. (APIServer pid=1833126) WARNING 08-07 13:30:05 [modelopt.py:1034] Detected ModelOpt NVFP4 checkpoint (quant_algo=NVFP4). Please note that the format is experimental and could change in future. (APIServer pid=1833126) WARNING 08-07 13:30:05 [modelopt.py:1034] Detected ModelOpt NVFP4 checkpoint (quant_algo=W4A16_NVFP4). Please note that the format is experimental and could change in future. (APIServer pid=1833126) WARNING 08-07 13:30:05 [modelopt.py:1707] Detected ModelOpt MXFP8 checkpoint. Please note that the format is experimental and could change in future. (APIServer pid=1833126) INFO 08-07 13:30:05 [vllm.py:1109] Asynchronous scheduling is enabled. (APIServer pid=1833126) INFO 08-07 13:30:05 [kernel.py:295] Final IR op priority after setting platform defaults: IrOpPriorityConfig(rms_norm=['native'], fused_add_rms_norm=['native']) tokenizer_config.json: 100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 16.7k/16.7k [00:00<00:00, 26.1MB/s] vocab.json: 100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 6.72M/6.72M [00:00<00:00, 32.2MB/s] tokenizer.json: downloading bytes: ████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 5.67MB, 532kB/s tokenizer.json: reconstructing file: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 12.8MB / 12.8MB, 1.22MB/s chat_template.jinja: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 7.76k/7.76k [00:00<00:00, 23.5MB/s] generation_config.json: 100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 202/202 [00:00<00:00, 998kB/s] video_preprocessor_config.json: 100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 385/385 [00:00<00:00, 1.41MB/s] (APIServer pid=1833126) [transformers] The `use_fast` parameter is deprecated and will be removed in a future version. Use `backend="torchvision"` instead of `use_fast=True`, or `backend="pil"` instead of `use_fast=False`. WARNING 08-07 13:30:34 [cuda.py:959] Detected different devices in the system: NVIDIA GeForce RTX 3060 Laptop GPU, NVIDIA GeForce RTX 5060 Ti. Please make sure to set `CUDA_DEVICE_ORDER=PCI_BUS_ID` to avoid unexpected behavior. (EngineCore pid=1833533) INFO 08-07 13:30:41 [core.py:116] Initializing a V1 LLM engine (v0.26.0) with config: model='nvidia/Qwen3.6-35B-A3B-NVFP4', speculative_config=None, tokenizer='nvidia/Qwen3.6-35B-A3B-NVFP4', skip_tokenizer_init=False, tokenizer_mode=auto, revision=None, tokenizer_revision=None, trust_remote_code=False, dtype=torch.bfloat16, max_seq_len=180000, download_dir=None, load_format=auto, tensor_parallel_size=1, pipeline_parallel_size=1, data_parallel_size=1, decode_context_parallel_size=1, dcp_comm_backend=ag_rs, disable_custom_all_reduce=False, quantization=modelopt_mixed, quantization_config=None, enforce_eager=False, enable_return_routed_experts=False, kv_cache_dtype=fp8, device_config=cuda, structured_outputs_config=StructuredOutputsConfig(backend='auto', disable_any_whitespace=False, disable_additional_properties=False, reasoning_parser='qwen3', reasoning_parser_plugin='', enable_in_reasoning=False), observability_config=ObservabilityConfig(show_hidden_metrics_for_version=None, otlp_traces_endpoint=None, collect_detailed_traces=None, kv_cache_metrics=False, kv_cache_metrics_sample=0.01, cudagraph_metrics=False, enable_layerwise_nvtx_tracing=False, enable_mfu_metrics=False, enable_mm_processor_stats=False, enable_logging_iteration_details=False, jit_monitor_mode='warn', jit_monitor_verbose=False), seed=0, served_model_name=nvidia/Qwen3.6-35B-A3B-NVFP4, enable_prefix_caching=False, enable_chunked_prefill=True, pooler_config=None, compilation_config={'mode': <CompilationMode.VLLM_COMPILE: 3>, 'debug_dump_path': None, 'cache_dir': '', 'compile_cache_save_format': 'binary', 'backend': 'inductor', 'custom_ops': ['none'], 'ir_enable_torch_wrap': True, 'splitting_ops': ['vllm::unified_attention_with_output', 'vllm::unified_mla_attention_with_output', 'vllm::mamba_mixer2', 'vllm::mamba_mixer', 'vllm::short_conv', 'vllm::linear_attention', 'vllm::plamo2_mamba_mixer', 'vllm::qwen_gdn_attention_core', 'vllm::gdn_attention_core_xpu', 'vllm::olmo_hybrid_gdn_full_forward', 'vllm::kda_attention', 'vllm::sparse_attn_indexer', 'vllm::rocm_aiter_sparse_attn_indexer', 'vllm::deepseek_v4_attention', 'vllm::hpc_rope_norm_forward', 'vllm::unified_kv_cache_update', 'vllm::unified_mla_kv_cache_update'], 'compile_mm_encoder': False, 'cudagraph_mm_encoder': False, 'encoder_cudagraph_token_budgets': [], 'encoder_cudagraph_max_vision_items_per_batch': 0, 'encoder_cudagraph_max_frames_per_batch': None, 'compile_sizes': [], 'compile_ranges_endpoints': [4096], 'inductor_compile_config': {'enable_auto_functionalized_v2': False, 'size_asserts': False, 'alignment_asserts': False, 'scalar_asserts': False, 'combo_kernels': True, 'benchmark_combo_kernel': True}, 'inductor_passes': {}, 'cudagraph_mode': <CUDAGraphMode.FULL_AND_PIECEWISE: (2, 1)>, 'cudagraph_num_of_warmups': 1, 'cudagraph_capture_sizes': [1, 2], 'cudagraph_copy_inputs': False, 'cudagraph_specialize_lora': True, 'use_inductor_graph_partition': False, 'pass_config': {'fuse_norm_quant': False, 'fuse_act_quant': False, 'fuse_attn_quant': False, 'enable_sp': False, 'fuse_gemm_comms': False, 'fuse_allreduce_rms': False, 'enable_qk_norm_rope_fusion': False, 'fuse_rope_kvcache_cat_mla': False, 'fuse_act_padding': False, 'fuse_qk_norm_rope_kvcache': False}, 'max_cudagraph_capture_size': 2, 'dynamic_shapes_config': {'type': <DynamicShapesType.BACKED: 'backed'>, 'evaluate_guards': False, 'assume_32_bit_indexing': False}, 'local_cache_dir': None, 'fast_moe_cold_start': False, 'static_all_moe_layers': []}, kernel_config=KernelConfig(ir_op_priority=IrOpPriorityConfig(rms_norm=['native'], fused_add_rms_norm=['native']), enable_flashinfer_autotune=True, enable_cutedsl_warmup=True, enable_bf16x3_router_gemm=False, moe_backend='auto', linear_backend='auto') (EngineCore pid=1833533) Warning: You are sending unauthenticated requests to the HF Hub. Please set a HF_TOKEN to enable higher rate limits and faster downloads. (EngineCore pid=1833533) INFO 08-07 13:30:45 [parallel_state.py:1615] world_size=1 rank=0 local_rank=0 distributed_init_method=tcp://10.3.2.87:52967 backend=nccl (EngineCore pid=1833533) INFO 08-07 13:30:45 [parallel_state.py:1946] rank 0 in world size 1 is assigned as DP rank 0, PP rank 0, PCP rank 0, TP rank 0, EP rank 0, EPLB rank N/A (EngineCore pid=1833533) Failed to get device capability: SM 12.x requires CUDA >= 12.9. (EngineCore pid=1833533) Failed to get device capability: SM 12.x requires CUDA >= 12.9. (EngineCore pid=1833533) INFO 08-07 13:30:49 [topk_topp_sampler.py:55] Using FlashInfer for top-p & top-k sampling. (EngineCore pid=1833533) [transformers] The `use_fast` parameter is deprecated and will be removed in a future version. Use `backend="torchvision"` instead of `use_fast=True`, or `backend="pil"` instead of `use_fast=False`. (EngineCore pid=1833533) INFO 08-07 13:31:05 [gpu_model_runner.py:5250] Starting to load model nvidia/Qwen3.6-35B-A3B-NVFP4... (EngineCore pid=1833533) INFO 08-07 13:31:05 [cuda.py:541] Using backend AttentionBackendEnum.FLASH_ATTN for vit attention (EngineCore pid=1833533) INFO 08-07 13:31:05 [mm_encoder_attention.py:373] Using AttentionBackendEnum.FLASH_ATTN for MMEncoderAttention. (EngineCore pid=1833533) INFO 08-07 13:31:05 [__init__.py:635] Selected MarlinFP8ScaledMMLinearKernel for ModelOptFp8LinearMethod (EngineCore pid=1833533) INFO 08-07 13:31:05 [qwen_gdn_linear_attn.py:150] Using Triton/FLA GDN prefill kernel (requested=auto, head_k_dim=128). (EngineCore pid=1833533) INFO 08-07 13:31:05 [nvfp4.py:285] Using 'MARLIN' NvFp4 MoE backend out of potential backends: ['FLASHINFER_TRTLLM', 'FLASHINFER_CUTEDSL', 'FLASHINFER_CUTEDSL_BATCHED', 'FLASHINFER_CUTLASS', 'VLLM_CUTLASS', 'MARLIN', 'HUMMING', 'EMULATION']. (EngineCore pid=1833533) INFO 08-07 13:31:05 [cuda.py:422] Using AttentionBackendEnum.FLASHINFER backend. (EngineCore pid=1833533) ERROR 08-07 13:31:07 [gpu_model_runner.py:5345] Failed to load model - not enough GPU memory. Try lowering --gpu-memory-utilization to free memory for weights, increasing --tensor-parallel-size, or using --quantization. See https://docs.vllm.ai/en/latest/configuration/conserving_memory/ for more tips. (original error: CUDA out of memory. Tried to allocate 256.00 MiB. GPU 0 has a total capacity of 15.52 GiB of which 68.62 MiB is free. Including non-PyTorch memory, this process has 15.44 GiB memory in use. Of the allocated memory 15.15 GiB is allocated by PyTorch, and 78.45 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://docs.pytorch.org/docs/stable/notes/cuda.html#optimizing-memory-usage-with-pytorch-cuda-alloc-conf))
  • Tuxedo17T
    Tuxedo17T Tuxedo17
    IA

    Chargement via :

    Available Repositories & VariantsStandard NVFP4 GGUF:

    • Get the single-file setup from Mudler Qwen3.6-27B-NVFP4-GGUF optimized for Blackwell hardware and paged-attention.
    • MTP (Multi-Token Prediction) Variants: Find improved scale-fitting layout options at Michaelw9999 Qwen3.6-27B NVFP4 MTP GGUF.
    • Alternative Community Quantizations: Check out CodeFault Unsloth Qwen3.6-27B NVFP4 GGUF for standard implementation wrappers.

    Hardware Requirements & Running Tips :

    • GPU Support: Native FP4 Tensor Cores (such as NVIDIA Blackwell architectures like the RTX 5090) provide maximum processing speed.
    • Fallback execution: Non-Blackwell hardware falls back to standard dequantization, which reduces execution speeds.
    • Backend: Use updated builds of llama.cpp with -fa 1 (paged/flash attention) enabled for optimal throughput
    6 0 0 Répondre
    Tuxedo17T Tuxedo17
    NVIDIA : https://huggingface.co/nvidia/Qwen3.6-35B-A3B-NVFP4 Exemple : models: # Premier modèle : Llama 3 8B - name: "meta-llama/Meta-Llama-3-8B-Instruct" command: > vllm serve meta-llama/Meta-Llama-3-8B-Instruct --port 8001 --gpu-memory-utilization 0.85 ready_url: "http://127.0.0" upstream_url: "http://127.0.0"
  • R
    R Remi
    IA

    Le PC en question :

    # inxi -F
System:
  Host: pcremi Kernel: 6.18.33.1-microsoft-standard-WSL2 arch: x86_64 bits: 64
  Console: pty pts/2 Distro: Ubuntu 24.04.4 LTS (Noble Numbat)
Machine:
  Message: No machine data: try newer kernel. Is dmidecode installed? Try -M
    --dmidecode.
CPU:
  Info: 6-core model: AMD Ryzen 5 7500X3D bits: 64 type: MT MCP cache:
    L2: 6 MiB
  Speed (MHz): avg: 3992 min/max: N/A cores: 1: 3992 2: 3992 3: 3992 4: 3992
    5: 3992 6: 3992 7: 3992 8: 3992 9: 3992 10: 3992 11: 3992 12: 3992
Graphics:
  Device-1: Microsoft Basic Render Driver driver: dxgkrnl v: 2.0.3
  Device-2: Microsoft Basic Render Driver driver: dxgkrnl v: 2.0.3
  Display: server: X.Org v: 24.1.6 driver: dri: swrast gpu: dxgkrnl,dxgkrnl
    resolution: 3840x2160~60Hz
  API: EGL v: 1.5 drivers: swrast platforms: wayland,x11,surfaceless,device
  API: OpenGL v: 4.6 vendor: microsoft mesa v: 24.0.9-0ubuntu0.3
    renderer: D3D12 (AMD Radeon RX 9060 XT)
Audio:
  Message: No device data found.
Network:
  Message: No PCI device data found.
  IF-ID-1: eth0 state: up speed: 10000 Mbps duplex: full
    mac: 
Drives:
  Local Storage: total: 1 TiB used: 1.93 GiB (0.2%)
  ID-1: /dev/sda model: Virtual Disk size: 356.9 MiB
  ID-2: /dev/sdb model: Virtual Disk size: 159.5 MiB
  ID-3: /dev/sdc model: Virtual Disk size: 4 GiB
  ID-4: /dev/sdd model: Virtual Disk size: 1024 GiB
Partition:
  ID-1: / size: 1006.85 GiB used: 1.93 GiB (0.2%) fs: ext4 dev: /dev/sdd
Swap:
  ID-1: swap-1 type: partition size: 4 GiB used: 0 KiB (0.0%) dev: /dev/sdc
Sensors:
  Src: lm-sensors+/sys Message: No sensor data found using /sys/class/hwmon
    or lm-sensors.
Info:
  Memory: total: 16 GiB note: est. available: 15.2 GiB used: 775.9 MiB (5.0%)
  Processes: 30 Uptime: 5m Shell: Bash inxi: 3.3.34
    29 0 1 Répondre
    R Remi
    Je fait un make install avant puis un nouveau test : root@pcremi:/workspace/llama.cpp/build/bin# ./llama-bench -m /models/Qwen3.6-27B-Q4_K_M.gguf ggml_cuda_init: found 1 ROCm devices (Total VRAM: 16304 MiB): Device 0: AMD Radeon RX 9060 XT, gfx1200 (0x1200), VMM: no, Wave Size: 32, VRAM: 16304 MiB | model | size | params | backend | ngl | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | --------------: | -------------------: | pid:3402 tid:0x757b61d04280 [CreateContext] fail 11 llama-bench: /home/remia/librocdxg/src/wddm/queue.cpp:267: wsl::thunk::ComputeQueue::ComputeQueue(wsl::thunk::WDDMDevice*, void*, uint64_t, std::atomic<long unsigned int>*, std::atomic<long unsigned int>*, volatile int64_t*, uint32_t, uint32_t, bool): Assertion `ret' failed. Aborted (core dumped) Pas mieux.