Commit graph

  • 014ed31c46
    Merge ad068e6ca8 into a71e89192e FatherlyFox 2026-10-21 22:51:08 -04:00
  • ad068e6ca8 data: community benchmark (amd-gpu) FatherlyFox 2026-10-21 22:51:06 -04:00
  • fc3bf2004a
    Merge ebdbcb79dd into a71e89192e kruegsche 2026-10-21 22:24:14 +02:00
  • ebdbcb79dd data: community benchmark (apple-m5-pro) kruegsche 2026-10-21 22:24:12 +02:00
  • 5eda1b2110
    Merge 04d0800198 into a71e89192e Baptiste Grosjean 2026-10-21 13:39:27 +00:00
  • 04d0800198 docs(bench): --tool-calls needs an OpenAI-compatible endpoint Baptiste Grosjean 2026-10-21 15:38:54 +02:00
  • 16bfb7d112
    Merge 19d9e46813 into a71e89192e Jordan Chin 2026-10-21 15:02:19 +02:00
  • 68e79802a5
    Merge 6226b16905 into a71e89192e qweflu 2026-10-21 15:02:19 +02:00
  • 78ff06da01
    Merge b6d0b1f711 into a71e89192e Rolf Larsson 2026-10-21 13:51:01 +01:00
  • 3d4aef5f52
    Merge 6b8fd6c2b9 into a71e89192e Steven Paul Sanderson II, MPH 2026-10-21 13:51:00 +01:00
  • 000dcc5b57
    Merge 4e4a9b8e4e into a71e89192e lorenzozane 2026-10-21 13:51:00 +01:00
  • 8fa4a3ea0f
    Merge 45b83975a2 into a71e89192e Alex Jones 2026-10-21 13:51:00 +01:00
  • 90946032a0
    Merge dbc3c796f2 into a71e89192e Aythami Estévez Olivas 2026-10-21 13:51:00 +01:00
  • c1070c7d13
    Merge 90bdd29b6a into a71e89192e kortasma 2026-10-21 13:51:00 +01:00
  • f5395d5017
    Merge c0a2d43ca6 into a71e89192e Ryan Ballantyne 2026-10-21 13:51:00 +01:00
  • ea28333df6
    Merge f2af43d69c into a71e89192e dependabot[bot] 2026-10-21 13:51:00 +01:00
  • 857fd1e5d7
    Merge 1eb89909db into a71e89192e github-actions[bot] 2026-10-21 13:51:00 +01:00
  • 9734f3b6cc
    Merge eb81a2177a into a71e89192e github-actions[bot] 2026-10-21 13:51:00 +01:00
  • ba5430e919
    Merge 78514c9553 into a71e89192e RyuShai 2026-10-21 13:51:00 +01:00
  • 07ea7b7457
    Merge 204b3a61e8 into a71e89192e yiheng-kkk 2026-10-21 13:50:59 +01:00
  • 947990c894
    Merge 534258452d into a71e89192e Baptiste Grosjean 2026-10-21 13:15:50 +02:00
  • 506abae298
    Merge 86f7e1286c into a71e89192e Baptiste Grosjean 2026-10-21 13:15:48 +02:00
  • f5e2f700ee
    Merge 178d14d4ae into a71e89192e take566 2026-10-21 10:41:59 +01:00
  • 569acb25b4
    Merge f2656ee0a3 into a71e89192e Ehsan-hash 2026-10-21 02:38:44 -07:00
  • f2656ee0a3 fix(tui): keep inference bench selection visible Ehsan-hash 2026-10-21 02:38:14 -07:00
  • f24976a2c4
    Merge 37729409bd into a71e89192e Jay Hemnani 2026-10-21 12:48:57 +05:30
  • ecd9ee497b
    Merge 971fd53534 into a71e89192e Jay Hemnani 2026-10-21 12:48:52 +05:30
  • 750daf9ff3
    Merge e98a01447d into a71e89192e Waldir Pimenta 2026-10-21 08:09:18 +01:00
  • 5a1a8aba5b
    Merge 86ca54b7b9 into a71e89192e Spicy Shark 2026-10-21 06:32:10 +00:00
  • 79f59d4c8e
    Merge 35a5db4145 into a71e89192e Spicy Shark 2026-10-21 06:32:07 +00:00
  • c48c802d6b
    Merge 65f32b3edd into a71e89192e Alex Jones 2026-10-21 06:26:57 +00:00
  • 65f32b3edd chore: update Cargo.lock release-please--branches--main github-actions[bot] 2026-10-21 06:26:54 +00:00
  • e5f34dcf4d chore(main): release 1.1.18 Alex Jones 2026-10-21 07:26:48 +01:00
  • a71e89192e
    docs: add WinGet installation instructions for Windows (#1175) main Ehsan-hash 2026-10-20 23:26:18 -07:00
  • 9c8a0603d0
    Merge 968d376d43 into bffd1b66b2 hide3803 2026-10-21 13:33:52 +09:00
  • 968d376d43 data: community benchmark (nvidia-rtx-pro-4000-blackwell-sff-edition) hide3803 2026-10-21 13:33:50 +09:00
  • ad4c90aac3 data: community benchmark (nvidia-rtx-pro-4000-blackwell-sff-edition) hide3803 2026-10-21 13:33:48 +09:00
  • 4a02e828d6 data: community benchmark (nvidia-rtx-pro-4000-blackwell-sff-edition) hide3803 2026-10-21 13:33:47 +09:00
  • 6779870873 docs: add WinGet installation instructions for Windows Ehsan-hash 2026-10-20 10:25:05 -07:00
  • 3b9b68608e fix(bench): --tool-calls refuses --all instead of ignoring it Baptiste Grosjean 2026-10-20 17:49:34 +02:00
  • 86f7e1286c test(bench): keep the ollama host test's result out of the user's store Baptiste Grosjean 2026-10-20 17:41:58 +02:00
  • e98a01447d data: community benchmark (intel-graphics) Waldir Pimenta 2026-10-20 16:30:46 +01:00
  • 3b051bca4a data: community benchmark (intel-graphics) Waldir Pimenta 2026-10-20 16:30:45 +01:00
  • ab6b3fe701 data: community benchmark (intel-graphics) Waldir Pimenta 2026-10-20 16:30:43 +01:00
  • b56c0554c9 data: community benchmark (intel-graphics) Waldir Pimenta 2026-10-20 16:30:42 +01:00
  • 42be1941db data: community benchmark (intel-graphics) Waldir Pimenta 2026-10-20 16:30:41 +01:00
  • 2ac0d4d53a data: community benchmark (intel-graphics) Waldir Pimenta 2026-10-20 16:30:39 +01:00
  • 534258452d fix(cli): refuse conflicting sizing flags split across the subcommand Baptiste Grosjean 2026-10-20 16:45:52 +02:00
  • 86ca54b7b9
    fix(plan): estimate F32 speed from 4 bytes per parameter spicysharkco 2026-10-20 09:10:36 +00:00
  • 35a5db4145
    fix(fit): size usable context from the active experts under MoE offload spicysharkco 2026-10-20 08:16:31 +00:00
  • bffd1b66b2
    fix(hardware): skip legacy Intel iGPUs with lspci controller names (#1148) Jimmy Liu Jia 2026-10-20 14:17:57 +08:00
  • 65ddd28f77
    fix(scraper): keep the stated context window for llama3 rope scaling (#1167) Jay Hemnani 2026-10-20 11:47:51 +05:30
  • 628184eb5c
    fix(claim): reject --quant labels it does not recognize (#1162) Jay Hemnani 2026-10-20 11:47:47 +05:30
  • 37729409bd fix(scraper): use the GGUF header count when the config estimate is lower Jay Hemnani 2026-10-20 00:59:41 +05:30
  • 6a0e233c3e fix(scraper): keep the stated context window for llama3 rope scaling Jay Hemnani 2026-10-20 01:02:20 +05:30
  • 7e53fe9043 fix(models): recognize the native ternary quant labels Jay Hemnani 2026-10-20 00:08:25 +05:30
  • 839b43e455 fix(bench): read OLLAMA_HOST like the Ollama provider does Baptiste Grosjean 2026-10-19 18:18:22 +02:00
  • 71b9d1b182 feat(bench): --tool-calls probes whether a model returns tool calls Baptiste Grosjean 2026-10-19 18:06:29 +02:00
  • 36e5371e6a fix(cli): accept the sizing flags after the subcommand Baptiste Grosjean 2026-10-19 18:06:28 +02:00
  • 646cbe2172 fix(claim): reject --quant labels it does not recognize Jay Hemnani 2026-10-19 20:03:19 +05:30
  • bba31622ce fix(hardware): cover legacy Intel generation names and lspci fallback Jimmy Liu Jia 2026-10-19 19:17:06 +08:00
  • b6d0b1f711 data: community benchmark (amd-radeon-rx-7900-xt) Rolf Larsson 2026-10-19 10:01:59 +02:00
  • 90900e60cb data: community benchmark (amd-radeon-rx-7900-xt) Rolf Larsson 2026-10-19 10:01:12 +02:00
  • 1ba341e40c data: community benchmark (amd-radeon-rx-7900-xt) Rolf Larsson 2026-10-19 10:00:21 +02:00
  • b42ce21510 data: community benchmark (amd-radeon-rx-7900-xt) Rolf Larsson 2026-10-19 09:59:55 +02:00
  • 74ec4c6bbc data: community benchmark (amd-radeon-rx-7900-xt) Rolf Larsson 2026-10-19 09:59:13 +02:00
  • 9450e3ba93 data: community benchmark (amd-radeon-rx-7900-xt) Rolf Larsson 2026-10-19 09:58:20 +02:00
  • a3fb6ce574 data: community benchmark (amd-radeon-rx-7900-xt) Rolf Larsson 2026-10-19 09:57:15 +02:00
  • 3e0da510cf data: community benchmark (amd-radeon-rx-7900-xt) Rolf Larsson 2026-10-19 09:56:08 +02:00
  • 36285a1621 data: community benchmark (amd-radeon-rx-7900-xt) Rolf Larsson 2026-10-19 09:42:44 +02:00
  • 9007a8706b data: community benchmark (amd-radeon-rx-7900-xt) Rolf Larsson 2026-10-19 09:42:08 +02:00
  • 83d9655129 data: community benchmark (amd-radeon-rx-7900-xt) Rolf Larsson 2026-10-19 09:41:47 +02:00
  • 3602913422 data: community benchmark (amd-radeon-rx-7900-xt) Rolf Larsson 2026-10-19 09:41:11 +02:00
  • f8e3e68fc0 data: community benchmark (amd-radeon-rx-7900-xt) Rolf Larsson 2026-10-19 09:40:38 +02:00
  • 6bf9392209 data: community benchmark (amd-radeon-rx-7900-xt) Rolf Larsson 2026-10-19 09:40:15 +02:00
  • 74a69896f6 data: community benchmark (amd-radeon-rx-7900-xt) Rolf Larsson 2026-10-19 09:39:18 +02:00
  • d79309c44d data: community benchmark (amd-radeon-rx-7900-xt) Rolf Larsson 2026-10-19 09:19:44 +02:00
  • fbde9aaf8d
    chore(deps): bump the cargo-minor-patch group with 4 updates (#1156) dependabot[bot] 2026-10-19 07:17:08 +01:00
  • 42f2395fde
    fix(cli): reject --memory and --ram values that do not parse (#1160) Jay Hemnani 2026-10-19 11:47:05 +05:30
  • eb81a2177a chore(models): Weekly model data refresh automated/weekly-model-update AlexsJones 2026-10-19 02:59:54 +00:00
  • 1eb89909db chore(data): Weekly benchmark cache refresh automated/weekly-benchmark-update AlexsJones 2026-10-19 02:11:04 +00:00
  • f2af43d69c
    chore(deps): bump the github-actions-minor-patch group with 4 updates dependabot/github_actions/github-actions-minor-patch-6aa41dcf00 dependabot[bot] 2026-10-19 00:55:43 +00:00
  • 6addfcd20b
    chore(deps): bump the cargo-minor-patch group with 4 updates dependabot[bot] 2026-10-19 00:55:14 +00:00
  • 0bc36b6248 fix(cli): reject --memory and --ram values that do not parse Jay Hemnani 2026-10-19 02:57:01 +05:30
  • c0a2d43ca6 data: community benchmark (radeon-680m) Ryan Ballantyne 2026-10-18 23:26:36 +10:00
  • 5095ace4cf data: community benchmark (radeon-680m) Ryan Ballantyne 2026-10-18 23:26:35 +10:00
  • 3c1e8fe25a data: community benchmark (radeon-680m) Ryan Ballantyne 2026-10-18 23:26:33 +10:00
  • 971fd53534 style: cargo fmt Jay Hemnani 2026-10-18 13:45:12 +05:30
  • 02f4d090e4 feat(hardware): detect and show the PCIe link of NVIDIA GPUs Jay Hemnani 2026-10-18 13:06:00 +05:30
  • e915091dd5
    fix(serve): only replace a stale socket at the --unix-socket path (#1152) Jay Hemnani 2026-10-18 11:56:06 +05:30
  • 22cc61a024 fix(serve): bound the stale-socket check so a busy server can't hang startup Jay Hemnani 2026-10-18 11:35:23 +05:30
  • 8ebfeb87aa fix(serve): keep the unix-only guards on unix_socket_router and its test Jay Hemnani 2026-10-18 11:23:50 +05:30
  • 90bdd29b6a data: community benchmark (radeon-rx-7900-xt-7900-xtx) kortasma 2026-10-17 12:06:04 -07:00
  • 40efc408be fix(serve): only replace a stale socket at the --unix-socket path Jay Hemnani 2026-10-18 00:24:23 +05:30
  • 2491243295
    fix(claim): size the MoE bandwidth floor from active parameters (#1150) Jay Hemnani 2026-10-17 11:45:27 +05:30
  • f6c95cc0e3 fix(claim): size the MoE bandwidth floor from active parameters Jay Hemnani 2026-10-16 20:57:48 +05:30
  • dbc3c796f2 data: community benchmark (nvidia-geforce-rtx-5060-ti) Aythami Estévez Olivas 2026-10-16 13:49:16 +02:00
  • 215bfc7b19 data: community benchmark (nvidia-geforce-rtx-5060-ti) Aythami Estévez Olivas 2026-10-16 13:15:29 +02:00
  • eb8db38351
    feat(cli): show backend-incompatible models on request (#1143) Jimmy Liu Jia 2026-10-16 00:00:37 +08:00
  • 0490df0fca
    fix(fit): size GPU-loaded MoE models at the quant they are rated at (#1141) Jay Hemnani 2026-10-15 21:30:31 +05:30