repo
stringclasses
15 values
fix_commit
stringlengths
40
40
buggy_commit
stringlengths
40
40
message
stringlengths
3
64.3k
files
listlengths
1
300
timestamp
timestamp[s]date
2013-03-13 20:45:00
2026-04-11 07:48:46
ggml-org/llama.cpp
f851fa5ab056c9bada48ad7208fe122fc0574e44
f1ac84119ccc8e72dafd9e9f8fc3b9399917ce11
fix: add openssl to nix dependencies (#21353) (#21355)
[ { "path": ".devops/nix/package.nix", "patch": "@@ -16,7 +16,7 @@\n rocmPackages,\n vulkan-headers,\n vulkan-loader,\n- curl,\n+ openssl,\n shaderc,\n useBlas ?\n builtins.all (x: !x) [\n@@ -160,7 +160,8 @@ effectiveStdenv.mkDerivation (finalAttrs: {\n ++ optionals useMpi [ mpi ]\n ++...
2026-04-03T09:21:07
huggingface/transformers
42c2ccb537135819858afdc5fbe8874ffb03a7dc
a4a176171c47979125025041adc4f8d201aec310
Fix peft conversion typo (#44053) fix
[ { "path": "src/transformers/integrations/peft.py", "patch": "@@ -261,7 +261,7 @@ def _build_peft_weight_mapping(\n source_patterns=new_source_patterns,\n target_patterns=new_target_patterns,\n distributed_operation=orig_conversion.distributed_opera...
2026-02-16T18:08:32
golang/go
1fe65836aca009343274443411f7120d70434ddb
6293a0683171e3a08e275884fddeaa2b57a861f6
cmd/go/testdata/script: skip EndSequence entries in build_trimpath_cgo The TestScript/build_trimpath_cgo test for cmd/go has been accessing a nil pointer when it tries to look up LineEntry.File.Name on a line entry with EndSequence set to true. The doc for EndSequence specifies that if EndSequence is set, only it and ...
[ { "path": "src/cmd/go/testdata/script/build_trimpath_cgo.txt", "patch": "@@ -121,6 +121,9 @@ func run(exePath string) ([]string, error) {\n \t\t\t\t}\n \t\t\t\treturn nil, err\n \t\t\t}\n+\t\t\tif le.EndSequence {\n+\t\t\t continue\n+\t\t\t}\n \t\t\tfiles[le.File.Name] = true\n \t\t}\n \t}", "additio...
2024-12-03T21:23:18
ollama/ollama
d75557747357bfb3afd441a0cc207ec944bd3a18
a2cc8571c5b2b8f77a8a5e2f65cb7aaa56482dc4
llm: Estimate projector memory correctly for Ollama engine The Llama engine always places vision projectors on the first GPU if one exists. However, the Ollama engine groups it with the output layer, which means the projector is only offloaded if all other layers are offloaded. The memory estimation code always assume...
[ { "path": "llm/memory.go", "patch": "@@ -85,8 +85,11 @@ func EstimateGPULayers(gpus []discover.GpuInfo, f *ggml.GGML, projectors []strin\n \tvar graphOffload uint64\n \n \t// Projectors loaded into GPU0 only\n-\tvar projectorWeights uint64\n-\tvar projectorGraph uint64\n+\tvar llamaEngineProjectorWeights ui...
2025-05-13T18:36:52
facebook/react
41a897797d92d1e5d43826da8340495b347f2f4f
5c8195aec7d16e19eed6d83f61f7d3acdd79db35
[Babel] Format errors for console logging
[ { "path": "compiler/forget/src/Babel/BabelPlugin.ts", "patch": "@@ -25,6 +25,7 @@ import {\n \n type BabelPluginPass = {\n opts: PluginOptions;\n+ filename: string | null;\n };\n \n function hasUseForgetDirective(directive: t.Directive): boolean {\n@@ -108,7 +109,7 @@ export default function ReactForgetB...
2023-05-15T13:45:26
ggml-org/llama.cpp
b069b10ab48f25ba119e59d0b8bf35d4f06e093f
0c58ba3365d2bc717b447b5d70e4d6be09ff3c40
vocab: fix Gemma4 tokenizer (#21343) * seems to work * fix case with new line Co-authored-by: sayap <sokann@gmail.com> * gemma 4: fix pre tok regex --------- Co-authored-by: Xuan Son Nguyen <son@huggingface.co> Co-authored-by: sayap <sokann@gmail.com>
[ { "path": "convert_hf_to_gguf.py", "patch": "@@ -7464,9 +7464,6 @@ def set_vocab(self):\n \n assert len(tokens) == vocab.vocab_size\n \n- # TODO @ngxson : there are some known (rare) issues with the tokenizer during development\n- # but I don't have time to dive into them right now;\n-...
2026-04-03T08:33:03
electron/electron
8c8642634dd8a1027ee9a4e2f41c97727ab66da1
afe0116d59f5abfb20d44af9525fb048809f5d73
fix: don't unmaximize on macOS if user set max bounds (#33480)
[ { "path": "shell/browser/native_window_mac.h", "patch": "@@ -267,6 +267,8 @@ class NativeWindowMac : public NativeWindow,\n // Maximizable window state; necessary for persistence through redraws.\n bool maximizable_ = true;\n \n+ bool user_set_bounds_maximized_ = false;\n+\n // Simple (pre-Lion) Full...
2022-03-30T22:29:02
golang/go
6293a0683171e3a08e275884fddeaa2b57a861f6
c3a706e1d2ea900fbaa715fb0100fb034abb1414
doc/next: introduce module tools Fixes #48429 Change-Id: Ie1954fc3bedd7d36cafaf3835b6c3a92326393f5 Reviewed-on: https://go-review.googlesource.com/c/go/+/632556 Reviewed-by: Michael Matloob <matloob@golang.org> LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com> Auto-Submit: Dmit...
[ { "path": "doc/next/3-tools.md", "patch": "@@ -2,7 +2,31 @@\n \n ### Go command {#go-command}\n \n+<!-- go.dev/issue/48429 -->\n+\n+Go modules can now track executable dependencies using `tool` directives in\n+go.mod. This removes the need for the previous workaround of adding tools as\n+blank imports to a ...
2024-12-01T05:13:52
huggingface/transformers
a4a176171c47979125025041adc4f8d201aec310
349e00c1a367ce263624e525038250625dcf20c7
[fix][wip] GlmMoeDsa: try implement DSA (#43912) * init Signed-off-by: JaredforReal <w13431838023@gmail.com> * format Signed-off-by: JaredforReal <w13431838023@gmail.com> * not indexer_rope_interleave Signed-off-by: JaredforReal <w13431838023@gmail.com> * set MLA rope interleave to False Signed-off-by: Jaredfor...
[ { "path": "src/transformers/models/glm_moe_dsa/configuration_glm_moe_dsa.py", "patch": "@@ -35,90 +35,91 @@ class GlmMoeDsaConfig(PreTrainedConfig):\n \n Args:\n vocab_size (`int`, *optional*, defaults to 154880):\n- Vocabulary size of the Deep model. Defines the number of different t...
2026-02-16T18:06:03
facebook/react
5c8195aec7d16e19eed6d83f61f7d3acdd79db35
dda3e61347fde65b62ed2b17ead463685cb35814
[prettier] Fix linting
[ { "path": "compiler/forget/src/ReactiveScopes/PropagateScopeDependencies.ts", "patch": "@@ -538,12 +538,15 @@ class PropagationVisitor extends ReactiveFunctionVisitor<Context> {\n id,\n scope: context.currentScope,\n });\n- } else if (value.kind === \"DeclareLocal\" || value.kind ==...
2023-05-15T13:43:39
ggml-org/llama.cpp
7c7d6ce5c7caad3714ad1cd37ef057dfe0f317d0
5208e2d5bae10fcbb3f3a21d0815441b2954374a
[HIP] Bump ROCm version to 7.2.1 (#21066) Bump ROCm version on Linux from 7.2 to 7.2.1 Add gfx1102 target Delete LLVM workaround since ROCm 7.2.1 has fix for ROCm 7.2 perf regression https://github.com/ROCm/rocm-systems/issues/2865 --------- Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com>
[ { "path": ".devops/rocm.Dockerfile", "patch": "@@ -1,8 +1,8 @@\n ARG UBUNTU_VERSION=24.04\n \n # This needs to generally match the container host's environment.\n-ARG ROCM_VERSION=7.2\n-ARG AMDGPU_VERSION=7.2\n+ARG ROCM_VERSION=7.2.1\n+ARG AMDGPU_VERSION=7.2.1\n \n # Target the ROCm build image\n ARG BASE_R...
2026-04-02T22:59:20
electron/electron
afe0116d59f5abfb20d44af9525fb048809f5d73
b711860d21da111270c3c59945a037d6d04eec83
fix: build when pdf component is disabled (#33513)
[ { "path": "BUILD.gn", "patch": "@@ -363,10 +363,6 @@ source_set(\"electron_lib\") {\n \"//components/network_session_configurator/common\",\n \"//components/omnibox/browser:buildflags\",\n \"//components/os_crypt\",\n- \"//components/pdf/browser\",\n- \"//components/pdf/browser:interceptor...
2022-03-30T18:55:38
huggingface/transformers
349e00c1a367ce263624e525038250625dcf20c7
2bd902dbc45ff87d6cd1986f8d2d40765bc34853
Fix early image processor return not raising error (#43981) * Fix early return not raising error * Fix the test instead * Revert other changes * Apply suggestion from @Rocketknight1
[ { "path": "tests/models/llava/test_image_processing_llava.py", "patch": "@@ -225,6 +225,11 @@ def pad_to_square_original(\n np.testing.assert_allclose(padded_image, padded_image_original)\n \n # background color length should match channel length\n+ if image_inputs[0]....
2026-02-16T16:40:39
ggml-org/llama.cpp
5208e2d5bae10fcbb3f3a21d0815441b2954374a
7992aa7c8e21ea2eb7a5e4802da56eec7b376036
fix: gemma 4 template (#21326)
[ { "path": "common/chat-auto-parser-generator.cpp", "patch": "@@ -7,11 +7,109 @@\n #include \"log.h\"\n #include \"nlohmann/json.hpp\"\n \n+#include <algorithm>\n #include <stdexcept>\n #include <string>\n \n using json = nlohmann::ordered_json;\n \n+namespace {\n+\n+// Gemma4-specific PEG builder extending ...
2026-04-02T21:31:02
golang/go
aa464fb953beabae11e94bffde320457f0904352
be297ba9b32a56630a655393142bb33bafd297df
doc/next: add release notes for Wasm changes Document wasmexport and WASI reactor/library mode. Also document that we now permit more types for wasmimport. Fixes #65199. Updates #66984. For #68545. Change-Id: Id26a8c9496650cd154daed679b82223af1848eea Reviewed-on: https://go-review.googlesource.com/c/go/+/634036 Revi...
[ { "path": "doc/next/7-ports.md", "patch": "@@ -14,5 +14,21 @@ Go 1.25 will require macOS 12 Monterey or later.\n \n ### WebAssembly {#wasm}\n \n+<!-- go.dev/issue/65199, CL 603055 -->\n+The `go:wasmexport` directive is added for Go programs to export functions to the WebAssembly host.\n+\n+On WebAssembly Sy...
2024-12-05T16:35:37
facebook/react
4cd7065665ea2cf33c306265c8d817904bb401ca
a389046a529c6d22ba5895dd7f5d4b0b8d17c345
Fix uSES hydration in strict mode (#26791) Previously, we'd call and use getSnapshot on the second render resulting in `Warning: Text content did not match. Server: "Nay!" Client: "Yay!"` and then `Error: Text content does not match server-rendered HTML.`. Fixes #26095. Closes #26113. Closes #25650. --------- ...
[ { "path": "packages/react-dom/src/__tests__/ReactDOMFizzServer-test.js", "patch": "@@ -20,6 +20,7 @@ let JSDOM;\n let Stream;\n let Scheduler;\n let React;\n+let ReactDOM;\n let ReactDOMClient;\n let ReactDOMFizzServer;\n let Suspense;\n@@ -73,6 +74,7 @@ describe('ReactDOMFizzServer', () => {\n \n Sched...
2023-05-12T21:18:03
huggingface/transformers
2bd902dbc45ff87d6cd1986f8d2d40765bc34853
2546978b80b90b02d7099bd0c3e8718a6bfac3a9
Update assignee for generate in bug report template (#44042) Joao is regrettably no longer with us :saluting_face: so we should really stop getting users to ping him! This PR makes @cyrilvallez responsible for `generate` issues outside of VLMs.
[ { "path": ".github/ISSUE_TEMPLATE/bug-report.yml", "patch": "@@ -44,7 +44,7 @@ body:\n \n Library:\n \n- - generate: @zucchini-nlp (visual-language models) or @gante (all others)\n+ - generate: @zucchini-nlp (visual-language models) or @cyrilvallez (all others)\n - contin...
2026-02-16T16:08:41
electron/electron
b711860d21da111270c3c59945a037d6d04eec83
929fc8bea5dd8f5fa64723cc0a805319e3340bb8
chore: bump chromium to 102.0.4971.0 (main) (#33454) * chore: bump chromium in DEPS to 102.0.4965.0 * chore: 3-way merge of chromium/printing.patch * chore: update patch shear in chromium/picture-in-picture.patch * chore: update patches * 3101519: Window Placement: Prototype fullscreen companion window sup...
[ { "path": ".circleci/build_config.yml", "patch": "@@ -59,7 +59,7 @@ executors:\n description: \"xcode version\"\n default: \"12.4.0\"\n type: enum\n- enum: [\"12.4.0\", \"13.2.1\"]\n+ enum: [\"12.4.0\", \"13.3.0\"]\n \n macos:\n xcode: << parameters.xcode >>\n...
2022-03-30T18:08:58
golang/go
be297ba9b32a56630a655393142bb33bafd297df
080466fffb47f420456b4a19129ffb271cae5aa6
cmd/go: drop -v from go help get Fixes #37301 Change-Id: I6d6ac818a73b907638f62d56bd5a7f00a6e6a5ba Reviewed-on: https://go-review.googlesource.com/c/go/+/632178 Auto-Submit: Dmitri Shuralyov <dmitshur@golang.org> Reviewed-by: Michael Knyszek <mknyszek@google.com> LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-projec...
[ { "path": "src/cmd/go/alldocs.go", "patch": "@@ -669,7 +669,7 @@\n //\n // Usage:\n //\n-//\tgo get [-t] [-u] [-v] [-tool] [build flags] [packages]\n+//\tgo get [-t] [-u] [-tool] [build flags] [packages]\n //\n // Get resolves its command-line arguments to packages at specific module versions,\n // updates ...
2024-11-28T16:23:53
facebook/react
dda3e61347fde65b62ed2b17ead463685cb35814
fe95b5df903903da6b372dcc2fd43f842ab3226d
[test cases] Add failing / bug tests for destructuring assignment
[ { "path": "compiler/forget/src/__tests__/fixtures/compiler/_bug.destructure-assignment-to-global.expect.md", "patch": "@@ -0,0 +1,29 @@\n+\n+## Input\n+\n+```javascript\n+function useFoo(props) {\n+ [x] = props;\n+ return { x };\n+}\n+\n+```\n+\n+## Code\n+\n+```javascript\n+import { unstable_useMemoCache...
2023-05-12T16:48:18
ggml-org/llama.cpp
7992aa7c8e21ea2eb7a5e4802da56eec7b376036
a1cfb645307edc61a89e41557f290f441043d3c2
tests : add unit test coverage for llama_tensor_get_type (#20112) * Add unit test coverage for llama_tensor_get_type * Fix merge conflicts, add more schemas * clang formatter changes * Trailing whitespace * Update name * Start rebase * Updating files with upstream changes prior to rebase * Changes needed from r...
[ { "path": "src/llama-ext.h", "patch": "@@ -1,12 +1,56 @@\n #pragma once\n \n-#include \"llama-context.h\"\n-#include \"ggml.h\"\n-#include \"stdint.h\"\n+#include \"llama.h\"\n+\n+#include <cstdint>\n \n // Reserve a new compute graph. It is valid until the next call to llama_graph_reserve.\n LLAMA_API stru...
2026-04-02T20:53:58
ollama/ollama
27da2cddc514208f4e23539e00485b880e3e2191
feb8923adada675b19dc1bc20f39ed6cfb0b99da
Fix lingering Q4_0 help reference (#10720)
[ { "path": "cmd/cmd.go", "patch": "@@ -1318,7 +1318,7 @@ func NewCLI() *cobra.Command {\n \t}\n \n \tcreateCmd.Flags().StringP(\"file\", \"f\", \"\", \"Name of the Modelfile (default \\\"Modelfile\\\"\")\n-\tcreateCmd.Flags().StringP(\"quantize\", \"q\", \"\", \"Quantize model to this level (e.g. q4_0)\")\n+...
2025-05-15T23:33:23
huggingface/transformers
2546978b80b90b02d7099bd0c3e8718a6bfac3a9
08c4ad95d4b3e995e3b61e50ffc4f3ba66642663
Add a dim check mechanism in Transpose and fix qwen3_vl_moe weight mapping (#44037) * start * test * betetr * fix * change name * cannot revert
[ { "path": "src/transformers/conversion_mapping.py", "patch": "@@ -141,17 +141,14 @@ def _build_checkpoint_conversion_mapping():\n ],\n \"qwen3_vl_moe\": [\n WeightConverter(\n- source_patterns=[\n- \"mlp.experts.*.gate_proj.weight\",\n- ...
2026-02-16T16:01:10
golang/go
080466fffb47f420456b4a19129ffb271cae5aa6
be0c4547d21e7e29ba062af7240b67e2263e5325
embed: document exclusions more explicitly Fixes #45197 Fixes #45744 Change-Id: Ic221da1b652f693c7aa37287932c0a8726f46765 Reviewed-on: https://go-review.googlesource.com/c/go/+/633815 Reviewed-by: Michael Knyszek <mknyszek@google.com> Reviewed-by: Ian Lance Taylor <iant@google.com> Auto-Submit: Ian Lance Taylor <iant...
[ { "path": "src/embed/embed.go", "patch": "@@ -90,7 +90,8 @@\n // depending on whether the package wants to make the data available to other packages.\n // It can only be used with variables at package scope, not with local variables.\n //\n-// Patterns must not match files outside the package's module, such...
2024-12-04T21:56:45
facebook/react
67a05d03e38b9837e27c9fe0a673557e63ff03c5
df12d7eac40c87bd5fdde0aa5a739bce9e7dce27
fix[devtools]: fixed duplicated backend activation with multiple renderers (#26807) ## Summary Initially reported in https://github.com/facebook/react/issues/26797. Was not able to reproduce the exact same problem, but found this case: 1. Open corresponding codepen from the issue in debug mode 2. Open components...
[ { "path": "packages/react-devtools-extensions/src/backend.js", "patch": "@@ -29,5 +29,6 @@ function setup(hook: ?DevToolsHook) {\n initBackend,\n setupNativeStyleEditor,\n });\n+\n hook.emit('devtools-backend-installed', COMPACT_VERSION_NAME);\n }", "additions": 1, "deletions": 0, "l...
2023-05-12T13:59:53
electron/electron
929fc8bea5dd8f5fa64723cc0a805319e3340bb8
a508dce1bfe000ad25a6d6320f12cd41ee1a3fd7
test: make test apps' name prefixed with electron-test- (#33511)
[ { "path": "spec-main/api-app-spec.ts", "patch": "@@ -90,9 +90,9 @@ describe('app module', () => {\n \n it('overrides the name', () => {\n expect(app.name).to.equal('Electron Test Main');\n- app.name = 'test-name';\n+ app.name = 'electron-test-name';\n \n- expect(app.name)....
2022-03-30T17:17:34
ggml-org/llama.cpp
a1cfb645307edc61a89e41557f290f441043d3c2
5803c8d11529b08844d740c03d8b6603608654b1
ggml-webgpu: add vectorized flash attention (#20709) * naive vectorized version * add vectorized flash attention * update vec version * remove unused path and shader * remove unused helper functions * add comments * remove pad path * ggml-webgpu: fix flash-attn vec nwg=1 path and tighten vec specialization * c...
[ { "path": "ggml/src/ggml-webgpu/ggml-webgpu-shader-lib.hpp", "patch": "@@ -95,6 +95,12 @@ struct ggml_webgpu_generic_shader_decisions {\n uint32_t wg_size = 0;\n };\n \n+struct ggml_webgpu_processed_shader {\n+ std::string wgsl;\n+ std::string variant;\n+ std::shared_ptr<voi...
2026-04-02T17:40:42
ollama/ollama
499ae7311fd26cb4e655ebea69712de3e242f629
ef202789fad6b8a8ab51f4d2ff5450067e3d1f65
ollamarunner: Base cached tokens on current prompt When we restore a sequence from the cache, we split the prompt into the already used tokens (stored in the cache) and new tokens that need to be processed. Currently, the references to the used tokens are coming from the stored previous sequence. However, even though...
[ { "path": "runner/llamarunner/cache.go", "patch": "@@ -104,8 +104,8 @@ func (c *InputCache) LoadCacheSlot(prompt []input, cachePrompt bool) (*InputCach\n \tslog.Debug(\"loading cache slot\", \"id\", slot.Id, \"cache\", len(slot.Inputs), \"prompt\", len(prompt),\n \t\t\"used\", numPast, \"remaining\", len(pr...
2025-05-09T23:51:47
huggingface/transformers
08c4ad95d4b3e995e3b61e50ffc4f3ba66642663
53f8a08290bf835c9891094352f9efd7da0ccece
fix(models): Unpack BitNet packed weights to fix CI failure (#43721) * fix: Unpack packed weights in quantizer post-loading hook (AutoBitLinear) * nit: Fix ci/circleci: check_code_quality * fix: Simply unpack BitNet weights during loading itself * fix: Address comments * review: Add tests * nit: Change the import...
[ { "path": "src/transformers/integrations/bitnet.py", "patch": "@@ -368,3 +368,38 @@ def replace_with_bitnet_linear(model, modules_to_not_convert: list[str] | None =\n )\n \n return model\n+\n+\n+class BitNetDeserialize:\n+ def __init__(self, hf_quantizer):\n+ self.hf_quantizer = hf_qua...
2026-02-16T15:38:18
electron/electron
a9296229c8038f667ae93a3acb245e2093983453
9c3b159b956d30a60052a4c6c81b297ad5329567
feat: add app.isHidden API for macOS (#32155) * feat: add app.isHidden API * Update docs/api/app.md Co-authored-by: Samuel Maddock <samuel.maddock@gmail.com> * fixed isHidden tests * Update docs/api/app.md Co-authored-by: John Kleinschmidt <jkleinsc@github.com> * Update spec-main/api-app-spec.ts C...
[ { "path": "docs/api/app.md", "patch": "@@ -606,6 +606,10 @@ You should seek to use the `steal` option as sparingly as possible.\n \n Hides all application windows without minimizing them.\n \n+### `app.isHidden()` _macOS_\n+\n+Returns `boolean` - `true` if the application—including all of its windows—is hid...
2022-03-30T01:34:07
golang/go
be0c4547d21e7e29ba062af7240b67e2263e5325
adfd81d8cdd9eaca4a5775abd174bc90ce313fdd
doc/next: add linker change release notes Document that we now generate GNU build ID or Mach-O UUID by default, and the related flags to disable or override it. Fixes #68678. Fixes #70586. For #68545. Change-Id: Ic79f3c05aab8a5032ba6983b8c72eeb4202bfe84 Reviewed-on: https://go-review.googlesource.com/c/go/+/633875 R...
[ { "path": "doc/next/5-toolchain.md", "patch": "@@ -10,6 +10,13 @@ whether directly or indirectly (through an alias type).\n \n ## Linker {#linker}\n \n+<!-- go.dev/issue/68678, go.dev/issue/68652, CL 618598, CL 618601 -->\n+The linker now generates a GNU build ID (the ELF `NT_GNU_BUILD_ID` note) on ELF plat...
2024-12-04T22:24:54
facebook/react
f5bdf462b2385add21a6d042c13a9b10919c2c9e
e649bf6429206a0f5bc865373648861849a6ff0e
Enable hooks validation on playground, make deguggable Enables hooks validation in playground. Also adds a tab to show the output of validation (in case it passes) with the inferred post dominator tree. We can use this to debug the dominator in case of false negatives. <img width="1724" alt="Screenshot 2023-05-11...
[ { "path": "compiler/forget/packages/playground/components/Editor/Output.tsx", "patch": "@@ -7,6 +7,11 @@\n \n import generate from \"@babel/generator\";\n import * as t from \"@babel/types\";\n+import {\n+ CodeIcon,\n+ DocumentAddIcon,\n+ InformationCircleIcon,\n+} from \"@heroicons/react/outline\";\n im...
2023-05-11T19:39:59
ggml-org/llama.cpp
5803c8d11529b08844d740c03d8b6603608654b1
63f8fe0ef4d134378fcdd67f50b05dff3cee179b
tests: allow exporting graph ops from HF file without downloading weights (#21182) * tests: allow exporting graph ops from HF file without downloading weights * use unique_ptr for llama_context in HF metadata case * fix missing non-required tensors falling back to type f32 * use unique pointers where possible * us...
[ { "path": "common/arg.cpp", "patch": "@@ -537,9 +537,11 @@ static bool common_params_parse_ex(int argc, char ** argv, common_params_context\n } catch (const std::exception & e) {\n LOG_WRN(\"HF cache migration failed: %s\\n\", e.what());\n }\n+ // export_graph_ops loads only metadata\n+ ...
2026-04-02T16:19:20
ollama/ollama
ef202789fad6b8a8ab51f4d2ff5450067e3d1f65
55760195e654992ca5f364aa191b24611b3b7531
fix pixel values padding (#10718) * panic if trying to pad 4d * fix pixel values padding
[ { "path": "ml/backend/ggml/ggml.go", "patch": "@@ -915,6 +915,8 @@ func (t *Tensor) RMSNorm(ctx ml.Context, w ml.Tensor, eps float32) ml.Tensor {\n func (t *Tensor) Pad(ctx ml.Context, shape ...int) ml.Tensor {\n \tif len(shape) != 4 {\n \t\tpanic(\"expected 4 dimensions\")\n+\t} else if shape[3] != 0 {\n+\...
2025-05-15T20:44:44
huggingface/transformers
53f8a08290bf835c9891094352f9efd7da0ccece
9a9231632eeb8be48f31db992b6f212ce34ab30b
SINQ quantization strategy integration (adapted for Transformers V5) (#43112) * sinq integration files * sinq integration update * sinq integration no lazy import * Tests for sinq integration * minor changes to sinq integration * sinq integration documentation added * small correction to sinq documentation * sm...
[ { "path": "docs/source/en/_toctree.yml", "patch": "@@ -235,6 +235,8 @@\n title: SpQR\n - local: quantization/vptq\n title: VPTQ\n+ - local: quantization/sinq\n+ title: SINQ\n - local: quantization/contribute\n title: Contribute\n title: Quantization", "additions": 2, "deletions...
2026-02-16T15:08:42
electron/electron
9c3b159b956d30a60052a4c6c81b297ad5329567
89c1abd1b325c8ac815f37521754e377120a4171
fix: on macOS show BrowserWindow on maximize if not currently shown (#32949)
[ { "path": "shell/browser/native_window_mac.mm", "patch": "@@ -602,13 +602,23 @@ void ViewDidMoveToSuperview(NSView* self, SEL _cmd) {\n }\n \n void NativeWindowMac::Maximize() {\n- if (IsMaximized())\n+ const bool is_visible = [window_ isVisible];\n+\n+ if (IsMaximized()) {\n+ if (!is_visible)\n+ ...
2022-03-29T23:08:34
golang/go
adfd81d8cdd9eaca4a5775abd174bc90ce313fdd
c46ba1f9efeb8efa6d8f513ec92f36ae1bfb8cd0
weak: fix subheading syntax in package doc Go doc comments only support one level of heading. Change-Id: I458f211bc7547a0676a59fdcb72f52ca480bf184 Reviewed-on: https://go-review.googlesource.com/c/go/+/633881 Auto-Submit: Michael Knyszek <mknyszek@google.com> Reviewed-by: Ian Lance Taylor <iant@google.com> LUCI-TryBo...
[ { "path": "src/weak/doc.go", "patch": "@@ -9,7 +9,7 @@ canonicalization maps (like the unique package), and for tying together\n the lifetimes of separate values (for example, through a map with weak\n keys).\n \n-## Advice\n+# Advice\n \n This package is intended to target niche use-cases like the unique\n...
2024-12-05T04:36:17
facebook/react
e649bf6429206a0f5bc865373648861849a6ff0e
d1e044b81ab311b74427ebc58fc541a53bc3c711
Fixtures from ESLint rule Incorporates the fixtures from eslint-plugin-react-hooks using a script, so that we can easily update them in the future. For each fixture we run the compiler with and without hooks validation first so that we know if the fixture is expected to pass — we have some false positives and fals...
[ { "path": "compiler/forget/scripts/build-react-hooks-fixures.js", "patch": "@@ -71,14 +71,15 @@ for (const fixture of fixtures) {\n let code = fixture.code;\n let prefix = \"\";\n if (error !== null) {\n- prefix = `todo.`;\n+ prefix = `todo.bail.`;\n code = `// @skip\\n// Unsupported input\\...
2023-05-11T19:39:58
ollama/ollama
55760195e654992ca5f364aa191b24611b3b7531
bd68d3ae50c67ba46ee94a584fa6d0386e4b8522
fix mllama conversion (#10716) cross attention Q and K projections needs to have their heads swapped, similar to non-cross attention Q and K tensors
[ { "path": "convert/convert_llama.go", "patch": "@@ -139,7 +139,8 @@ func (p *llamaModel) Tensors(ts []Tensor) []*ggml.Tensor {\n \t}\n \n \tfor _, t := range ts {\n-\t\tif strings.HasSuffix(t.Name(), \"attn_q.weight\") || strings.HasSuffix(t.Name(), \"attn_k.weight\") {\n+\t\tif strings.HasSuffix(t.Name(), ...
2025-05-15T19:15:01
huggingface/transformers
360d72623e71adc475377f71a976612eed513395
8bb26307a61f97084b193e111da4190650507663
[Deespeed] fix WeightConverter.convert() use (#43926) * fix WeightConverter.convert() use * fix tests * directly checks no keys were missed * Apply repo consistency fixes --------- Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
[ { "path": "src/transformers/integrations/deepspeed.py", "patch": "@@ -347,17 +347,12 @@ def _apply_weight_conversions_to_state_dict(model, state_dict, weight_mapping):\n # Sort keys to ensure consistent ordering (important for MoE conversions)\n # Iterate over sorted keys and pop from state_dict to ...
2026-02-16T14:06:32
ggml-org/llama.cpp
63f8fe0ef4d134378fcdd67f50b05dff3cee179b
223373742bc1bd48e37b22192d1302f54d6f14bc
model, mtmd: fix gguf conversion for audio/vision mmproj (#21309) * fix gguf conversion for audio/vision mmproj * fix test
[ { "path": "common/chat-auto-parser-generator.cpp", "patch": "@@ -169,6 +169,8 @@ common_peg_parser analyze_tools::build_parser(parser_build_context & ctx) const\n return build_tool_parser_tag_json(ctx);\n case tool_format::TAG_WITH_TAGGED:\n return build_tool_parser_tag_tagge...
2026-04-02T15:10:32
golang/go
32ff485c7cd0d8e31a89723b56f6324ff77ac448
61232096579d8801cb3513c84a97c141d50ad242
math/bits: update reference to debruijn paper The old link no longer works. Fixes #70684 Change-Id: I8711ef7d5721bf20ef83f5192dd0d1f73dda6ce1 Reviewed-on: https://go-review.googlesource.com/c/go/+/633775 Auto-Submit: Ian Lance Taylor <iant@golang.org> Reviewed-by: Keith Randall <khr@golang.org> Reviewed-by: Michael ...
[ { "path": "src/math/bits/bits.go", "patch": "@@ -38,7 +38,7 @@ func LeadingZeros64(x uint64) int { return 64 - Len64(x) }\n \n // --- TrailingZeros ---\n \n-// See http://supertech.csail.mit.edu/papers/debruijn.pdf\n+// See http://keithandkatie.com/keith/papers/debruijn.html\n const deBruijn32 = 0x077CB531\...
2024-12-04T21:50:30
electron/electron
4c988a5a24a6e1dd81875c2ec3b2b7ed47d5c96a
94498b923e407ab7e6ac094ae829c89744ebf65c
docs: fix type of 'value' param in setUserDefaults (#33481)
[ { "path": "docs/api/system-preferences.md", "patch": "@@ -177,11 +177,11 @@ Some popular `key` and `type`s are:\n * `NSPreferredWebServices`: `dictionary`\n * `NSUserDictionaryReplacementItems`: `array`\n \n-### `systemPreferences.setUserDefault(key, type, value)` _macOS_\n+### `systemPreferences.setUserDef...
2022-03-29T21:47:06
ollama/ollama
ff80718e9c9a08dd10759cdee5c81db366e38368
0aa8b371ddd24a2d0ce859903a9284e9544f5c78
fix crash in old clients with quantization progress (#10710) Older clients assumed the digest was at least 19 characters long so increase the size of the dummy digest to avoid array out of bounds crashes.
[ { "path": "server/create.go", "patch": "@@ -430,7 +430,7 @@ func quantizeLayer(layer *layerGGML, quantizeType string, fn func(resp api.Progr\n \tfnWrap := func(n uint64) {\n \t\tdone := doneBytes.Add(n)\n \t\tprogress := float32(done) / float32(totalBytes)\n-\t\tfn(api.ProgressResponse{Status: fmt.Sprintf(\...
2025-05-14T21:54:18
facebook/react
d1e044b81ab311b74427ebc58fc541a53bc3c711
a2169d894a380c184036ea020a8e149857eb3941
[deps] Add DeclareLocal to ReactiveScope decls --- Try to fix bug from #1589: > If a declaration for an immutable identifier (i.e. one that is not later re-assigned, since undefined is a primitive) is sandwiched between mutations, we currently do not record it as an output or hoist it out of the reactive scope....
[ { "path": "compiler/forget/src/ReactiveScopes/PropagateScopeDependencies.ts", "patch": "@@ -537,6 +537,15 @@ class PropagationVisitor extends ReactiveFunctionVisitor<Context> {\n id,\n scope: context.currentScope,\n });\n+ } else if (value.kind === \"DeclareLocal\") {\n+ // Som...
2023-05-11T20:24:37
golang/go
61232096579d8801cb3513c84a97c141d50ad242
6f42fe914169ef237423fb07e9c8321884b58c23
cmd/go/internal/cache: always check error from stat in markUsed markUsed was not checking that the error from os.Stat was nil before trying to access the FileInfo entry returned by it. Instead, always check the error and return false if it's non-nil (usually because the file does not exist). This can happen if an inde...
[ { "path": "src/cmd/go/internal/cache/cache.go", "patch": "@@ -314,8 +314,8 @@ func GetMmap(c Cache, id ActionID) ([]byte, Entry, error) {\n // OutputFile returns the name of the cache file storing output with the given OutputID.\n func (c *DiskCache) OutputFile(out OutputID) string {\n \tfile := c.fileName(...
2024-12-03T20:09:41
ggml-org/llama.cpp
e15efe007dc1c0d79afa347190dba91de3bd659b
6137c325a16073c8bf68a52396a815006ccaa9a9
Relax prefill parser to allow space. (#21240) * Relax prefill parser to allow space. * Move changes from prefix() to parser generation * Only allow spaces if we're not having a pure content parser next
[ { "path": "common/chat-auto-parser-generator.cpp", "patch": "@@ -100,19 +100,22 @@ common_peg_arena autoparser::build_parser(const generation_params & inputs) cons\n \n bool has_tools = inputs.tools.is_array() && !inputs.tools.empty();\n bool has_response_format = inputs.json_schem...
2026-04-02T09:29:11
electron/electron
d1ea62c3e835abc20e30e64d76823edfa5cbe75c
8ea0631b823aea7f01b1eb9e410ea74dbb857556
fix: getting focused window with destroyed webContents (#33404) * fix: getting focused window with destroyed webContents * fix: add extra safeguards
[ { "path": "lib/browser/api/browser-window.ts", "patch": "@@ -72,7 +72,10 @@ BrowserWindow.getAllWindows = () => {\n \n BrowserWindow.getFocusedWindow = () => {\n for (const window of BrowserWindow.getAllWindows()) {\n- if (window.isFocused() || window.isDevToolsFocused()) return window;\n+ const has...
2022-03-29T16:22:58
ollama/ollama
0478d440f0ba62202bc4b98043ae4a7d0b85e4ba
8cc33f4c2ba9347b3de3b5fe197d486df741d3e4
Fixed over vram allcation dure to small initial layer sizes. Co-authored-by: Tej Kiran <kiran.tej@amd.com> Co-authored-by: Michael Yang <mxyng@pm.me> Co-authored-by: Tej Kiran <itej89@gmailcom>
[ { "path": "llm/memory.go", "patch": "@@ -1,9 +1,12 @@\n package llm\n \n import (\n+\t\"cmp\"\n \t\"fmt\"\n \t\"log/slog\"\n+\t\"maps\"\n \t\"os\"\n+\t\"slices\"\n \t\"strconv\"\n \t\"strings\"\n \n@@ -120,12 +123,10 @@ func EstimateGPULayers(gpus []discover.GpuInfo, f *ggml.GGML, projectors []strin\n \t}\n...
2025-05-13T23:42:39
facebook/react
a2169d894a380c184036ea020a8e149857eb3941
77c535ce25575e0679d7df8d40afe4dcce1d3c1f
[test case] Test for DeclareLocal instruction --- Record incorrect output. If a declaration for an immutable identifier (i.e. one that is not later re-assigned, since `undefined` is a primitive) is sandwiched between mutations, we currently do not record it as an output or hoist it out of the reactive scope.
[ { "path": "compiler/forget/src/__tests__/fixtures/compiler/_bug.uninitialized-declaration-in-reactive-scope.expect.md", "patch": "@@ -0,0 +1,40 @@\n+\n+## Input\n+\n+```javascript\n+function Component(props) {\n+ let x = mutate();\n+ let y;\n+ foo(x);\n+ return [y, x];\n+}\n+\n+```\n+\n+## Code\n+\n+```...
2023-05-11T20:24:36
ggml-org/llama.cpp
4888137b1736b706e39806025d24e4ca342f1e4a
fbd441c37933550c1e3365dc84dd73232334c15d
sycl : fix llama_kv_cache hang when kv_cache is huge: 5GB (#21283)
[ { "path": "ggml/src/ggml-sycl/ggml-sycl.cpp", "patch": "@@ -569,9 +569,15 @@ static void ggml_backend_sycl_buffer_clear(ggml_backend_buffer_t buffer,\n SYCL_CHECK(\n CHECK_TRY_ERROR(dpct::get_current_device().queues_wait_and_throw()));\n \n- SYCL_CHECK(CHECK_TRY_ERROR((*stream)\n- ...
2026-04-02T07:08:32
golang/go
6f42fe914169ef237423fb07e9c8321884b58c23
d241ea8d5cc5dce6c32c69d1a6303622874d5fd3
bufio: document Reader must be created with New functions Fixes #37347 Change-Id: Id642dbe7e52933f8d6cf7116c3c27ddb74efba9d Reviewed-on: https://go-review.googlesource.com/c/go/+/633315 LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com> Reviewed-by: Robert Griesemer <gri@google....
[ { "path": "src/bufio/bufio.go", "patch": "@@ -29,6 +29,9 @@ var (\n // Buffered input.\n \n // Reader implements buffering for an io.Reader object.\n+// A new Reader is created by calling [NewReader] or [NewReaderSize];\n+// alternatively the zero value of a Reader may be used after calling [Reset]\n+// on ...
2024-12-03T21:19:24
ollama/ollama
8cc33f4c2ba9347b3de3b5fe197d486df741d3e4
f46df4e5d2e964ccfd0f23f9377240b6d9897ed8
llama: fix memory leak for grammar (#10696)
[ { "path": "llama/sampling_ext.cpp", "patch": "@@ -114,6 +114,9 @@ void grammar_free(struct llama_grammar *g) {\n if (g->vocab != nullptr) {\n delete g->vocab;\n }\n+ if (g->o_vocab != nullptr) {\n+ delete g->o_vocab;\n+ }\n llama_grammar_free_...
2025-05-13T22:39:27
electron/electron
f0c22a770d131acb6117f6d07d86fb9b8b66476c
bf3d0e225760a0026a1b58ac4f4d29a83b0aad20
chore: fix moveAbove desktopCapturer spec (#33471)
[ { "path": "spec-main/api-desktop-capturer-spec.ts", "patch": "@@ -1,7 +1,8 @@\n import { expect } from 'chai';\n import { screen, desktopCapturer, BrowserWindow } from 'electron/main';\n+import { delay, ifdescribe, ifit } from './spec-helpers';\n import { emittedOnce } from './events-helpers';\n-import { if...
2022-03-29T14:05:08
ggml-org/llama.cpp
fbd441c37933550c1e3365dc84dd73232334c15d
c30e012253dd9e322c8e3424f808a5c74ecc46bf
hexagon : add cumsum op support (#21246) * hexagon : add cumsum op support * hexagon: enable dma for cumsum op * Fix line-ending --------- Co-authored-by: Max Krasnyansky <maxk@qti.qualcomm.com>
[ { "path": "ggml/src/ggml-hexagon/ggml-hexagon.cpp", "patch": "@@ -2231,6 +2231,22 @@ static bool ggml_hexagon_supported_ssm_conv(const struct ggml_hexagon_session *\n return true;\n }\n \n+static bool ggml_hexagon_supported_cumsum(const struct ggml_hexagon_session * sess, const struct ggml_tensor * op) ...
2026-04-02T00:44:02
golang/go
bc0406333f686a754c2d787f93b143c27721925d
2a0aeb12345e4ca64315830ffe87e57e1a828570
sync: document RWMutex locks cannot be upgraded / downgraded Fixes #38859 Change-Id: I9f4f6fa4cb529aaadfb3812e3a5c0da982a95f68 Reviewed-on: https://go-review.googlesource.com/c/go/+/633415 Auto-Submit: Ian Lance Taylor <iant@google.com> Reviewed-by: Michael Knyszek <mknyszek@google.com> LUCI-TryBot-Result: Go LUCI <g...
[ { "path": "src/sync/rwmutex.go", "patch": "@@ -24,6 +24,8 @@ import (\n // the writer has acquired (and released) the lock, to ensure that\n // the lock eventually becomes available to the writer.\n // Note that this prohibits recursive read-locking.\n+// A [RWMutex.RLock] cannot be upgraded into a [RWMutex...
2024-12-03T22:41:38
facebook/react
77c535ce25575e0679d7df8d40afe4dcce1d3c1f
5e80821aa247362a5a9cf3e97754183cc0884e01
[snap] Make watch mode go fast(er) While optimizing per @josephsavona's suggestions in #1592, I noticed that we were clearing quite a few require cache entries. As of this PR, `Object.keys(require.cache)` holds - 1258 entries total - 67 files compiled from Forget source code (this is what `ts.createWatchCompi...
[ { "path": "compiler/forget/packages/snap/src/compiler-worker.ts", "patch": "@@ -11,24 +11,14 @@ import { exists } from \"./utils\";\n \n const originalConsoleError = console.error;\n \n-// Subpaths to ignore when clearing the require cache\n-const ignoredRequireSubpaths: Array<string> = [\n- // compiler wo...
2023-05-11T18:58:51
ollama/ollama
f46df4e5d2e964ccfd0f23f9377240b6d9897ed8
c6bcdc4223c50071b59a19c42cc54ec9932f696f
llama: fix defrag patch to defragment when no slots are available (#10695)
[ { "path": "llama/llama.cpp/src/llama-context.cpp", "patch": "@@ -950,9 +950,12 @@ int llama_context::decode(llama_batch & inp_batch) {\n \n // find KV slot\n if (!kv_self->find_slot(ubatch)) {\n- LLAMA_LOG_WARN(\"%s: failed to find KV cache slot for ubatch of size %d\\n\", __func_...
2025-05-13T21:02:08
huggingface/transformers
3955dc76bf6ab1513835e016a7ec4552492130b8
33d9b320967d3a47781fbfed418764edd319f2fa
Improve use of torch.is_autocast_enabled (#43930) * Improve use of torch.is_autocast_enabled Signed-off-by: Yuanyuan Chen <cyyever@outlook.com> * Fix documentation examples Signed-off-by: Yuanyuan Chen <cyyever@outlook.com> --------- Signed-off-by: Yuanyuan Chen <cyyever@outlook.com>
[ { "path": "docs/source/en/debugging.md", "patch": "@@ -288,8 +288,9 @@ One solution is to go back a few steps before the values started growing too lar\n import torch\n \n def forward(self, hidden_states):\n- if torch.is_autocast_enabled():\n- with torch.cuda.amp.autocast(enabled=False):\n+ dev...
2026-02-16T13:54:59
electron/electron
37a94d9857a611e38d799cb36790e872a5a722c6
4aeeb64d300555dacb6c3d17989266c3fd0bab57
ci: fixup esbuild on macos (#33486) * ci: fixup esbuild on macos * ci: call update_depot_tools right after clone * when all else fails, use sed
[ { "path": ".circleci/build_config.yml", "patch": "@@ -470,6 +470,8 @@ step-fix-sync: &step-fix-sync\n \n # Fix esbuild (wrong binary)\n echo 'infra/3pp/tools/esbuild/${platform}' `gclient getdep --deps-file=src/third_party/devtools-frontend/src/DEPS -r 'third_party/esbuild:infra/3pp/tools/es...
2022-03-29T03:57:11
ggml-org/llama.cpp
95a6ebabb277c4cc18247e7bc2a5502133caca63
12dbf1da9558524b315bc47fae976fe90ec3a95e
opencl: fix leak in Adreno q8_0 path (#21212)
[ { "path": "ggml/src/ggml-opencl/ggml-opencl.cpp", "patch": "@@ -9612,6 +9612,9 @@ static void ggml_cl_mul_mat_q8_0_f32_adreno(ggml_backend_t backend, const ggml_t\n cl_mem B_image1d;\n cl_mem B_sub_buffer;\n cl_mem S_image1d;\n+ // for B transpose\n+ ...
2026-04-01T19:54:58
golang/go
2a0aeb12345e4ca64315830ffe87e57e1a828570
953a1bec48662c57fb40f50a2a4f56f033d842c6
doc/next: minor grammar fix in release notes For #66626. For #68545. Change-Id: I4fe21bdaa9580eb087c849de72e98f6145121dd0 Reviewed-on: https://go-review.googlesource.com/c/go/+/633282 Reviewed-by: Robert Griesemer <gri@google.com> Auto-Submit: Robert Griesemer <gri@google.com> Reviewed-by: Alan Donovan <adonovan@goog...
[ { "path": "doc/next/6-stdlib/99-minor/go/types/66626.md", "patch": "@@ -1,6 +1,6 @@\n \n All `go/types` data structures that expose sequences using a pair of\n-methods such as `Len() int` and `At(int) T` now also methods that\n+methods such as `Len() int` and `At(int) T` now also have methods that\n return ...
2024-12-03T21:55:47
facebook/react
7cd98ef2bcbc10f164f778bade86a4daeb821011
b5810163e913e8c95a7a0a6dee039bc102e3c987
Fix nightly job to publish to "canary" channel (#26799) When I was renaming the next channel to canary, I updated the `publish_preleases` workflow correctly, but I skipped over `publish_preleases_nightly`. Oops.
[ { "path": ".circleci/config.yml", "patch": "@@ -646,7 +646,7 @@ workflows:\n name: Publish to Canary channel\n commit_sha: << pipeline.git.revision >>\n release_channel: stable\n- dist_tag: \"next\"\n+ dist_tag: \"canary,next\"\n - publish_prerelease:\n ...
2023-05-10T02:27:35
ollama/ollama
4b903f088aa8d14404e5650d42db8c15530803d5
c7f4ae7b9c8976b4d50c59eb87e9582ea9c5c82f
llama: fix crash on snowflake embedding model (#10690)
[ { "path": "llama/llama.cpp/src/llama-vocab.cpp", "patch": "@@ -1469,8 +1469,6 @@ void llama_vocab::impl::load(llama_model_loader & ml, const LLM_KV & kv) {\n const int precompiled_charsmap_keyidx = gguf_find_key(ctx, kv(LLM_KV_TOKENIZER_PRECOMPILED_CHARSMAP).c_str());\n if (precompil...
2025-05-13T20:11:11
ggml-org/llama.cpp
12dbf1da9558524b315bc47fae976fe90ec3a95e
86221cf6dace86f47d896a38e0de652db4aa81a8
server: Bypass API Key validation for WebUI static bundle assets (#21269) * fix: Bypass API Key validation for static bundle assets * refactor: All bypassed routes in `public_endpoints` * test: Update static assets API Key test
[ { "path": "tools/server/server-http.cpp", "patch": "@@ -143,16 +143,20 @@ bool server_http_context::init(const common_params & params) {\n \"/v1/health\",\n \"/models\",\n \"/v1/models\",\n- \"/api/tags\"\n+ \"/api/tags\",\n+ \"/\",\n+ ...
2026-04-01T19:32:15
huggingface/transformers
b1141006cd2284aa19b2661d2575b5b26acd17e8
b768d8b157e227c7c84e57014f5183d461556579
docs: fix typos across documentation files (#43993)
[ { "path": "docs/source/en/glossary.md", "patch": "@@ -89,7 +89,7 @@ See [causal language modeling](#causal-language-modeling) and [decoder models](#\n \n ### backbone\n \n-The backbone is the network (embeddings and layers) that outputs the raw hidden states or features. It is usually connected to a [head](...
2026-02-16T13:41:17
electron/electron
4aeeb64d300555dacb6c3d17989266c3fd0bab57
594dc7e24a71bff68b2b6e69801e881a910a5604
fix: crash when WindowButtonsProxy references cleared NSWindow (#33069) * resets WindowButtonsProxy on window delete on macOS * fixes reset
[ { "path": "shell/browser/native_window_mac.mm", "patch": "@@ -328,6 +328,8 @@ void ViewDidMoveToSuperview(NSView* self, SEL _cmd) {\n [](NativeWindowMac* window) {\n if (window->window_)\n window->window_ = nil;\n+ if (window->buttons_proxy_)\n+ window->buttons_proxy_...
2022-03-29T03:02:44
golang/go
953a1bec48662c57fb40f50a2a4f56f033d842c6
4daf7922f38cd7900fd531d0c113f3c1f3024e87
runtime: avoid defer on system stack in synctestidle_c Fixes #70661 Change-Id: I58a465cfb1cd16709ffbb072eca0997569540074 Reviewed-on: https://go-review.googlesource.com/c/go/+/633281 Auto-Submit: Damien Neil <dneil@google.com> Reviewed-by: Michael Pratt <mpratt@google.com> LUCI-TryBot-Result: Go LUCI <golang-scoped@l...
[ { "path": "src/runtime/synctest.go", "patch": "@@ -217,13 +217,15 @@ func synctestRun(f func()) {\n \n func synctestidle_c(gp *g, _ unsafe.Pointer) bool {\n \tlock(&gp.syncGroup.mu)\n-\tdefer unlock(&gp.syncGroup.mu)\n+\tcanIdle := true\n \tif gp.syncGroup.running == 0 && gp.syncGroup.active == 1 {\n \t\t//...
2024-12-03T21:18:38
facebook/react
f3d4d1691819c58c6f9e42b113750685ab647dd0
04cee98166d4b495b29454f4e2a774788442ecd0
[lower] Bail out on computed lvalues in destructure --- Currently, we're silently producing incorrect output - ```js // source function Component(props) { const computedKey = props.key; const { [computedKey]: x } = props.val; return x; } // compiles to function Component(props) { const { compute...
[ { "path": "compiler/forget/src/HIR/BuildHIR.ts", "patch": "@@ -2586,6 +2586,14 @@ function lowerAssignment(\n });\n continue;\n }\n+ if (property.node.computed) {\n+ builder.errors.push({\n+ reason: `(BuildHIR::lowerAssignment) Handle comput...
2023-05-09T18:38:15
ollama/ollama
526b2ed10296cc3d1ae89121eedcbbbe257741a3
a7240c6d636836f0bca01790038d7194f519604b
fix vocabulary (#10679)
[ { "path": "model/models/gemma2/model.go", "patch": "@@ -45,6 +45,8 @@ func New(c fs.Config) (model.Model, error) {\n \t\t\t\tTypes: c.Ints(\"tokenizer.ggml.token_type\"),\n \t\t\t\tBOS: int32(c.Uint(\"tokenizer.ggml.bos_token_id\")),\n \t\t\t\tEOS: int32(c.Uint(\"tokenizer.ggml.eos_token_id\")),\n+\t...
2025-05-13T00:29:46
electron/electron
594dc7e24a71bff68b2b6e69801e881a910a5604
1c5bbba5cf2ef82839e46778b44d7737e50a6f06
chore: update node types version (#33452) * chore: update node types version * update express types to solve type conflict * one more yarn.lock type bump * update another types package to fix incompatible global declarations * remove incompatible type magicks * update our ambient types to match the node...
[ { "path": "lib/common/webpack-provider.ts", "patch": "@@ -7,7 +7,7 @@\n \n // Rip global off of window (which is also global) so that webpack doesn't\n // auto replace it with a looped reference to this file\n-const _global = typeof globalThis !== 'undefined' ? globalThis.global : (self as any || window as ...
2022-03-28T21:37:35
golang/go
4daf7922f38cd7900fd531d0c113f3c1f3024e87
bdedc5c1d7facb2e6c803e0a319734e60025cbf5
cmd/link/internal/ld: work around FreeBSD 14.1 cmd/cgo test link failure On FreeBSD 14.1 we fail to link against C code with internal linking. The symptom is apparently undefined symbols, but explicitly pointing the linker at compiler-rt for -libgcc fixes the issue. This looks a lot like the workaround on OpenBSD, but...
[ { "path": "src/cmd/link/internal/ld/lib.go", "patch": "@@ -632,6 +632,13 @@ func (ctxt *Link) loadlib() {\n \t\t\tif *flagLibGCC == \"\" {\n \t\t\t\t*flagLibGCC = ctxt.findLibPathCmd(\"--print-libgcc-file-name\", \"libgcc\")\n \t\t\t}\n+\t\t\tif runtime.GOOS == \"freebsd\" && strings.HasPrefix(filepath.Base...
2024-12-02T19:10:37
facebook/react
04cee98166d4b495b29454f4e2a774788442ecd0
5de02b881b9a0c30e3bc5be352e54f4404ad72fa
[snap] patch bug in clearing require cache Snap currently has a bug in which the require cache is not correctly cleared when running in filter mode (#tests < 2 * #workers). - We're currently clearing all entries in the require cache of worker threads, including `jest-worker` and `snap/dist/...` files - jest-wor...
[ { "path": "compiler/forget/packages/snap/src/compiler-worker.ts", "patch": "@@ -11,10 +11,26 @@ import { exists } from \"./utils\";\n \n const originalConsoleError = console.error;\n \n+// Subpaths to ignore when clearing the require cache\n+const ignoredRequireSubpaths: Array<string> = [\n+ // compiler wo...
2023-05-11T17:53:06
huggingface/transformers
b768d8b157e227c7c84e57014f5183d461556579
737f1777d90aaf7ec94e8c72ac135b876f25a576
Add Voxtral Realtime (#43769) * draft structure * intermediary ckpt * matching generate * conv cache draft * cache update * update * use appropirate text model * add processor * streaming -> realtime * nit * handle max tokens * num_delay_tokens * nit * add tests * nit * test update * handle generator i...
[ { "path": "docs/source/en/_toctree.yml", "patch": "@@ -1281,6 +1281,8 @@\n title: VisualBERT\n - local: model_doc/voxtral\n title: Voxtral\n+ - local: model_doc/voxtral_realtime\n+ title: VoxtralRealtime\n - local: model_doc/xclip\n title: X-CLIP\n title...
2026-02-16T13:32:05
ollama/ollama
9d6df9080502adcb6f25950e3d829ab05ec8cfc8
0cefd46f23ed60d5b41a90e6b6a1864e4214da3b
Follow up to #10363 (#10647) The quantization PR didn't block all unsupported file types, which this PR fixes. It also updates the API docs to reflect the now reduced set of supported types.
[ { "path": "docs/api.md", "patch": "@@ -19,7 +19,7 @@\n \n ### Model names\n \n-Model names follow a `model:tag` format, where `model` can have an optional namespace such as `example/model`. Some examples are `orca-mini:3b-q4_1` and `llama3:70b`. The tag is optional and, if not provided, will default to `lat...
2025-05-12T22:23:31
golang/go
a8be834912fae7e1f3f5cfc39e6a15f8782efc4d
485ed2fa5b5e0b7067ef72a0f4bdc9ca12b77ed7
cmd/compile: fix a premature-deallocation of state in loopreschedchecks lastMems is free-ed before it's actually used. This results in the pass breaking: multiple mem phis will present in the instrumented ssa, and essentially break tighten pass(https://cs.opensource.google/go/go/+/master:src/cmd/compile/internal/ssa/t...
[ { "path": "src/cmd/compile/internal/ssa/loopreschedchecks.go", "patch": "@@ -71,6 +71,7 @@ func insertLoopReschedChecks(f *Func) {\n \t}\n \n \tlastMems := findLastMems(f)\n+\tdefer f.Cache.freeValueSlice(lastMems)\n \n \tidom := f.Idom()\n \tpo := f.postorder()\n@@ -406,7 +407,6 @@ func findLastMems(f *Fun...
2024-11-30T04:15:07
electron/electron
c4e3a1aad3ae3fc2be4fc8efa64b17abddb7b621
3c30b59c3e64c39ae5b6dffc293d13efc537d6db
docs: Use Node's URL parser in the 5th security recommendation (#33463) Rule 13 recommends using Node's URL parser for handling url inputs. At the moment, this is not being followed in the code example for rule 5, which falls back on checking that the url ends with a '/'. If this was forgotten when a user copies th...
[ { "path": "docs/tutorial/security.md", "patch": "@@ -279,19 +279,20 @@ security-conscious developers might want to assume the very opposite.\n \n ```js title='main.js (Main Process)'\n const { session } = require('electron')\n+const URL = require('url').URL\n \n session\n .fromPartition('some-partition')\...
2022-03-28T18:25:44
facebook/react
5de02b881b9a0c30e3bc5be352e54f4404ad72fa
50c3bd9d98906b8544a23e6efda31b235071cfb9
Support unused conditional/logical expressions assigned to variable I noticed some instances of this error when running forget on an internal product. I previously fixed the case if a logical/conditional used only for side effects (not assigned to a variable) but the new cases were assigned to an unused variable. ...
[ { "path": "compiler/forget/src/ReactiveScopes/BuildReactiveFunction.ts", "patch": "@@ -6,7 +6,6 @@\n */\n \n import invariant from \"invariant\";\n-import { CompilerError } from \"../CompilerError\";\n import {\n BasicBlock,\n BlockId,\n@@ -826,12 +825,6 @@ class Driver {\n testBlock.terminal...
2023-05-09T17:58:51
huggingface/transformers
8ead2903d5ea3799cc3bb43670fddccac1d80f23
d4bea33f493f11ee775df9e820fb88916afc9a88
stable grouped_mm API (#43977) * use stable api * keep torch 2.9 support * fix * don't pass bias * last * keep private * revert * fix dtype mismatch from mixed precision
[ { "path": "docs/source/en/experts_interface.md", "patch": "@@ -19,11 +19,11 @@ All Mixture-of-Experts (MoE) implementations perform the same high-level computa\n \n The [`ExpertsInterface`] provides optimized experts backends. It decouples the experts implementation from the model code to simplify experimen...
2026-02-16T11:09:30
golang/go
485ed2fa5b5e0b7067ef72a0f4bdc9ca12b77ed7
106c804751992ccbba94c1dbdea4845a96659aa5
crypto/rsa: return error if keygen random source is broken Fixes #70643 Change-Id: I47c76500bb2e79b0d1dc096651eb45885f6888b6 Reviewed-on: https://go-review.googlesource.com/c/go/+/632896 Reviewed-by: Russ Cox <rsc@golang.org> Reviewed-by: Roland Shoemaker <roland@golang.org> LUCI-TryBot-Result: Go LUCI <golang-scoped...
[ { "path": "src/crypto/internal/fips140/rsa/keygen.go", "patch": "@@ -45,6 +45,10 @@ func GenerateKey(rand io.Reader, bits int) (*PrivateKey, error) {\n \t\t\treturn nil, err\n \t\t}\n \n+\t\tif Q.Nat().ExpandFor(P).Equal(P.Nat()) == 1 {\n+\t\t\treturn nil, errors.New(\"rsa: generated p == q, random source i...
2024-12-02T20:04:15
ollama/ollama
0d6e35d3c67cf37de1c425d178c71d7351083013
6e9a7a256856bf1119c992cf7da39c05276f386c
fix: stream accumulator exits early (#10593) the stream accumulator exits as soon as it sees `api.ProgressResponse(status="success")` which isn't strictly correctly since some requests may have multiple successes, e.g. `/api/create` when the source model needs to be pulled.
[ { "path": "server/routes.go", "patch": "@@ -1341,31 +1341,29 @@ func Serve(ln net.Listener) error {\n \n func waitForStream(c *gin.Context, ch chan any) {\n \tc.Header(\"Content-Type\", \"application/json\")\n+\tvar latest api.ProgressResponse\n \tfor resp := range ch {\n \t\tswitch r := resp.(type) {\n \t\...
2025-05-08T20:17:30
electron/electron
cdf2b3f4e48799c3a6ab91d0d70b5aaaf3a44c0a
1153a5ce5a34b553466b0da563831ff59cc6d328
fix: calling of X11 functions when running under Wayland (#33355) * fix: don't call X11 functions in file dialog and message box * refactor: remove unused GtkUiPlatform declaration * fix: set gtk darktheme only when running under X11 * fix: replace X11 window state watcher with implementation using ozone *...
[ { "path": "BUILD.gn", "patch": "@@ -563,7 +563,6 @@ source_set(\"electron_lib\") {\n defines += [\n # Disable warnings for g_settings_list_schemas.\n \"GLIB_DISABLE_DEPRECATION_WARNINGS\",\n- \"USE_X11=1\",\n ]\n \n sources += [", "additions": 0, "deletions": 1, "lan...
2022-03-28T16:48:50
facebook/react
50c3bd9d98906b8544a23e6efda31b235071cfb9
90109b3572ed3089afd862fd92830f5ce4f447ad
Fix capturedRefs collection for lambdas When we calculate the dependencies of a FunctionExpression we were only adding new items if the binding identifier had not been seen yet. That is correct for `capturedIds` since its the set of identifiers, but incorrect for `capturedRefs` since its an array of all the distin...
[ { "path": "compiler/forget/src/HIR/BuildHIR.ts", "patch": "@@ -2753,8 +2753,9 @@ function gatherCapturedDeps(\n fn: NodePath<t.FunctionExpression | t.ArrowFunctionExpression>,\n componentScope: Scope\n ): { identifiers: t.Identifier[]; refs: Place[] } {\n- const capturedIds: Set<t.Identifier> = new Set...
2023-05-09T17:38:56
ggml-org/llama.cpp
8710e5f9b9bd7246608808ccd3626bde8abf6ff9
1d6d4cf7a5361046f778414c5b1f5ecbc07eeb77
hexagon: improve RMS_NORM and DIV accuracy (#21251) * hexagon-rms_norm: fix RMS_NORM for non-aligned tensor sizes Co-authored-by: Krishna Sridhar <srsr@qti.qualcomm.com> * hexagon-div: perform DIV in fp16 domain for lower dsp archs --------- Co-authored-by: Krishna Sridhar <srsr@qti.qualcomm.com>
[ { "path": "ggml/src/ggml-hexagon/htp/hvx-div.h", "patch": "@@ -16,8 +16,10 @@\n \n #if __HVX_ARCH__ < 79\n #define HVX_OP_MUL_F32(a, b) Q6_Vsf_equals_Vqf32(Q6_Vqf32_vmpy_VsfVsf(a, b))\n+#define HVX_OP_MUL_F16(a, b) Q6_Vhf_equals_Wqf32(Q6_Wqf32_vmpy_VhfVhf(a, b))\n #else\n #define HVX_OP_MUL_F32(a, b) Q6_Vsf...
2026-04-01T15:43:08
huggingface/transformers
d4bea33f493f11ee775df9e820fb88916afc9a88
0dda1ffbf39c8dabeb4511af86745fb74b10e5aa
Remove `other_workflow_run_ids` for `issue_comment` in `utils/notification_service.py` (#44036) fix Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>
[ { "path": "utils/notification_service.py", "patch": "@@ -1511,18 +1511,6 @@ def pop_default(l: list[Any], i: int, default: Any) -> Any:\n token=os.environ[\"ACCESS_REPO_INFO_TOKEN\"], workflow_id=other_workflow_id, commit_sha=ci_sha\n )\n other_workflow_run_ids.append...
2026-02-16T10:21:54
golang/go
106c804751992ccbba94c1dbdea4845a96659aa5
ab59a4f985d3272cacee94e12f55584bad1aa573
crypto/rsa: fix keys with p < q Updates #70643 Change-Id: I4aee8373dbddf774564902b3957c6eba11d15fc9 Reviewed-on: https://go-review.googlesource.com/c/go/+/632955 Reviewed-by: Roland Shoemaker <roland@golang.org> Auto-Submit: Filippo Valsorda <filippo@golang.org> Reviewed-by: Russ Cox <rsc@golang.org> LUCI-TryBot-Resu...
[ { "path": "src/crypto/internal/fips140/rsa/rsa.go", "patch": "@@ -265,8 +265,12 @@ func checkPrivateKey(priv *PrivateKey) error {\n \t}\n \n \t// Check that qInv * q ≡ 1 mod p.\n-\tone := q.Nat().Mul(priv.qInv, p)\n-\tif one.IsOne() != 1 {\n+\tqP, err := bigmod.NewNat().SetOverflowingBytes(q.Nat().Bytes(q),...
2024-12-02T19:08:12
ollama/ollama
5e380c3b42741541d01cea0c821f4e01aa4e432e
392de84031e71cbd97ffe19b89ccf6cfeed9c7b3
sched: fix race leading to orphaned runners (#10599) If a model is loading, and the request context is canceled during the load by a client closing the connection, and another request is inbound for the same model with a different configuration (context size, etc.) thus requiring a reload, two unload events can be in ...
[ { "path": "llm/server.go", "patch": "@@ -1010,17 +1010,17 @@ func (s *llmServer) Close() error {\n \ts.llamaModelLock.Unlock()\n \n \tif s.cmd != nil {\n-\t\tslog.Debug(\"stopping llama server\")\n+\t\tslog.Debug(\"stopping llama server\", \"pid\", s.Pid())\n \t\tif err := s.cmd.Process.Kill(); err != nil {...
2025-05-07T16:38:17
electron/electron
1153a5ce5a34b553466b0da563831ff59cc6d328
9e45a1cd51397c9c3eecb3988877dee75ff0e283
fix: BrowserView background color in webContents (#33435) * chore: fix BrowserView background color in webContents * disable screen capture test on linux * spec: fix platform failure condition
[ { "path": "shell/browser/api/electron_api_browser_view.cc", "patch": "@@ -6,10 +6,13 @@\n \n #include <vector>\n \n+#include \"content/browser/renderer_host/render_widget_host_view_base.h\" // nogncheck\n+#include \"content/public/browser/render_widget_host_view.h\"\n #include \"shell/browser/api/electron_...
2022-03-28T16:47:08
facebook/react
fda1f0b902b527089fe5ae7b3aa573c633166ec9
7ac5e9a602347f3b7d26c60a549c483d3bc88bbf
Flow upgrade to 0.205.1 (#26796) Just a small upgrade to keep us current and remove unused suppressions (probably fixed by some upgrade since). - `*` is no longer allowed and has been an alias for `any` for a while now.
[ { "path": "package.json", "patch": "@@ -65,8 +65,8 @@\n \"eslint-plugin-react-internal\": \"link:./scripts/eslint-rules\",\n \"fbjs-scripts\": \"^3.0.1\",\n \"filesize\": \"^6.0.1\",\n- \"flow-bin\": \"^0.202.0\",\n- \"flow-remove-types\": \"^2.202.0\",\n+ \"flow-bin\": \"^0.205.1\",\n+...
2023-05-09T14:45:50
ggml-org/llama.cpp
1d6d4cf7a5361046f778414c5b1f5ecbc07eeb77
744c0c7310aad90e99a29c5739e4ee317fb6a748
fix: tool call parsing for LFM2 and LFM2.5 models (#21242) * fix: tool call parsing for LFM2 and LFM2.5 models' * refactor: add test / break out lfm2 and lfm2.5 parsing logic
[ { "path": "common/chat.cpp", "patch": "@@ -1274,11 +1274,12 @@ static common_chat_params common_chat_params_init_kimi_k2(const common_chat_temp\n return data;\n }\n \n-// LFM2 format:\n-// - Reasoning: <think>{reasoning}</think> (optional, only if enable_thinking is true)\n-// - Content: text after reas...
2026-04-01T14:22:44
huggingface/transformers
44412f6e74114b8157fd1c168cfc6f462a284844
8008e6c83e1467dbe0ae3c81d19b29c17f4ff456
Fix TypeError in dot_natural_key when state_dict keys have mixed types at same position (#43966) * Fix TypeError in dot_natural_key when state_dict keys have mixed types When a model has sub-models with different naming conventions (e.g. 'model.layers.26.self_attn...' vs 'desc_model.roberta.encoder...'), dot_natural_...
[ { "path": "src/transformers/core_model_loading.py", "patch": "@@ -812,12 +812,17 @@ def _job():\n \n \n def dot_natural_key(s: str):\n- parts = s.split(\".\")\n- for i, p in enumerate(parts):\n- # whole-segment digits -> int; otherwise leave as str\n+ \"\"\"Sort key for state-dict names: spl...
2026-02-13T17:39:51
electron/electron
9e45a1cd51397c9c3eecb3988877dee75ff0e283
d4a34fb175858b91f50b9dd7a125be3c7c26aca1
docs: Update release dates for E19 and fix typos (#33464)
[ { "path": "docs/tutorial/electron-timelines.md", "patch": "@@ -22,23 +22,23 @@ check out our [Electron Versioning](./electron-versioning.md) doc.\n | 12.0.0 | -- | 2020-Nov-19 | 2021-Mar-02 | M89 | v14.16 | 🚫 |\n | 13.0.0 | -- | 2021-Mar-04 | 2021-May-25 | M91 | v14.16 | 🚫 |\n | 14.0.0 | -- | 2021-May-27 ...
2022-03-28T16:30:05
facebook/react
597e70d69f8a2c0023d0a001c819fed3c2c2ac3a
8920fc0b2f82f8922c082c271503e3c156ed85cc
Improve eslint derived fixtures
[ { "path": "compiler/forget/scripts/build-react-hooks-fixures.js", "patch": "@@ -16,6 +16,8 @@ const path = require(\"path\");\n const prettier = require(\"prettier\");\n const prettierConfigPath = require.resolve(\"../.prettierrc\");\n const process = require(\"process\");\n+const { createHash } = require(\...
2023-05-08T21:10:07
huggingface/transformers
0dfdd6247414d40a44ea28da1bf331a40f11b178
f73a4db3a0bcf6523e9bfdaaf4afe81dffba4da8
Fix sync gradient (#43919) * fix sync gradient * Fix * style * fix
[ { "path": "src/transformers/trainer.py", "patch": "@@ -218,6 +218,7 @@\n DataLoaderConfiguration,\n DistributedDataParallelKwargs,\n DistributedType,\n+ GradientAccumulationPlugin,\n load_fsdp_model,\n load_fsdp_optimizer,\n release_memory,\n@@ -1764,14...
2026-02-13T14:38:07
ollama/ollama
af31ccefc00ff7e98839dd0066c8ddb7a44a7ccb
fa393554b927f154145488c852297a2330cb5f13
fix data race in WriteGGUF (#10598) err in the go routine should not be shared with the outer scope
[ { "path": "fs/ggml/gguf.go", "patch": "@@ -569,7 +569,7 @@ func WriteGGUF(f *os.File, kv KV, ts []*Tensor) error {\n \t\tt := t\n \t\tw := io.NewOffsetWriter(f, offset+int64(t.Offset))\n \t\tg.Go(func() error {\n-\t\t\t_, err = t.WriteTo(w)\n+\t\t\t_, err := t.WriteTo(w)\n \t\t\treturn err\n \t\t})\n \t}", ...
2025-05-07T00:36:38
ggml-org/llama.cpp
0356e33aafcc1cb409910244482fac1ec8bafe9f
6422036fcb32d72e3aca93dcf97fdfeeca20c27d
scripts: add function call test script (#21234) * scripts: add function call test script * add reasoning_content * fix lint
[ { "path": "scripts/server-test-function-call.py", "patch": "@@ -0,0 +1,1135 @@\n+#!/usr/bin/env python3\n+\"\"\"\n+Test tool calling capability via chat completions endpoint.\n+\n+Each test case contains:\n+ - tools: list of tool definitions (OpenAI-compatible)\n+ - messages: initial conversation messages...
2026-04-01T13:31:58
facebook/react
8920fc0b2f82f8922c082c271503e3c156ed85cc
7dedc6cc724d168a81e0e2d7e9a943a0bb31a78e
Dont count throw as an exit node (React semantics) React will retry or abort components that throw (depending on a few conditions), so from React's perspective a `throw` statement is not a normal exit node. Thus the Rules of Hooks really have a caveat: the set of hooks that are called _in an execution that returns...
[ { "path": "compiler/forget/src/HIR/Dominator.ts", "patch": "@@ -113,10 +113,9 @@ class Dominator<T> {\n */\n get(id: T): T | null {\n const dominator = this.#nodes.get(id);\n- invariant(\n- dominator !== undefined,\n- `Called on invalid node identifier '${id}'`\n- );\n+ if (domin...
2023-05-08T20:58:46
electron/electron
b03d6dfba91733f8fed50da38f7a37b7c6be2f18
f60ff18b1488b5ba28d930d22924816452a638e4
fix: non-client windows messages on legacy widget host (again) (#33438)
[ { "path": "patches/chromium/fix_non-client_mouse_tracking_and_message_bubbling_on_windows.patch", "patch": "@@ -13,7 +13,7 @@ messages in the legacy window handle layer.\n These conditions are regularly hit with WCO-enabled windows on Windows.\n \n diff --git a/content/browser/renderer_host/legacy_render_wi...
2022-03-25T10:12:09
huggingface/transformers
f73a4db3a0bcf6523e9bfdaaf4afe81dffba4da8
3e4b8d63ef4b3e04add27827abbf30d6a8640f30
Inclusion of process_group in the gather_full_tensor function in tensor_parallel.py (#43932) * Bug fix on gather_full_tensor for upstreaming. * Made correct usage of process_group in all_gather now.
[ { "path": "src/transformers/integrations/tensor_parallel.py", "patch": "@@ -1156,14 +1156,16 @@ def gather_full_tensor(local_tensor: torch.Tensor, shard_dim: int, device_mesh)\n The full reconstructed tensor (same on all ranks)\n \"\"\"\n world_size = device_mesh.size()\n+ # In case of TP...
2026-02-13T14:20:47
ggml-org/llama.cpp
e1cb817483ebda4e3ebc30fd07f4292c654f4339
88d5f8ffc398f9d6dbe529e3f0f0c739eabeafa7
memory: respect unified KV cache in hybrid memory for eval tasks (#21224) The hybrid memory paths (`llama-memory-hybrid.cpp` and `llama-memory-hybrid-iswa.cpp`) always used sequential equal split, ignoring the unified KV cache flag. This caused hellaswag, winogrande, and multiple-choice evaluations to fail on hybrid m...
[ { "path": "src/llama-memory-hybrid-iswa.cpp", "patch": "@@ -73,9 +73,9 @@ llama_memory_context_ptr llama_memory_hybrid_iswa::init_batch(llama_batch_allocr\n // if all tokens are output, split by sequence\n ubatch = balloc.split_seq(n_ubatch);\n } else {\n- ...
2026-04-01T09:50:17