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 | 36dafba5c476297261692bfb24c49ec657030c62 | 69e0ecef06bfef7d35c91f791b78651d4a3422e3 | llama: fix llama-model-saver (#20503)
* llama : add fd-based model loading via llama_model_load_from_fd
* llama : address review feedback for fd-based model loading
* llama : use FILE pointer instead of fd in public API
* llama : use FILE pointer consistently, address review feedback
* fixup
* fix tensor names
*... | [
{
"path": "ggml/include/gguf.h",
"patch": "@@ -77,6 +77,7 @@ extern \"C\" {\n };\n \n GGML_API struct gguf_context * gguf_init_empty(void);\n+ GGML_API struct gguf_context * gguf_init_from_file_ptr(FILE * file, struct gguf_init_params params);\n GGML_API struct gguf_context * gguf_init_from_f... | 2026-03-25T10:53:16 |
facebook/react | e9abc41ea3f0170659292ce1151a015300105999 | db378d39f3091e4b9302e74b1c6651112d45888b | Fix evaluation order for JSX element tags
When lowering a JSX element we were correctly lowering to a temporary in all but
one case: the common case of an identifier. That is fine in practice but breaks
in the presence of the tag identifier being reassigned in the props/children.
This PR fixes to always lower the ... | [
{
"path": "compiler/forget/src/HIR/BuildHIR.ts",
"patch": "@@ -27,14 +27,14 @@ import {\n InstructionKind,\n InstructionValue,\n JsxAttribute,\n- makeInstructionId,\n ObjectPattern,\n ObjectProperty,\n Place,\n ReturnTerminal,\n SourceLocation,\n SpreadPattern,\n ThrowTerminal,\n+ make... | 2023-04-20T23:48:34 |
golang/go | 6a7733af45493ddc1ededd2d0545004102c811a2 | a86ea80197ed3bf0f276638a9ce079cbd2071d83 | crypto/pbkdf2: init package
This commit imports the x/crypto/pbkdf2 package as described in the
linked proposal. The code is unchanged with the exception of a few
small updates to reflect feedback from the proposal comment period:
* the Key function is made generic over a hash.Hash
* the h function is moved to be the... | [
{
"path": "api/next/69488.txt",
"patch": "@@ -0,0 +1 @@\n+pkg crypto/pbkdf2, func Key[$0 hash.Hash](func() $0, string, []uint8, int, int) ([]uint8, error) #69488",
"additions": 1,
"deletions": 0,
"language": "Plain Text"
},
{
"path": "doc/next/6-stdlib/2-pbkdf2.md",
"patch": "@@ -0,0... | 2024-11-14T18:38:14 |
electron/electron | 268cd31e38f4d8b0ba81de27c73b7b1424ec9157 | 5b2d3910c1749f5faaeaec15d1051a2582c32596 | fix: DCHECK when calling app.exit() (#33035) | [
{
"path": "shell/browser/window_list.cc",
"patch": "@@ -109,7 +109,7 @@ void WindowList::DestroyAllWindows() {\n ConvertToWeakPtrVector(GetInstance()->windows_);\n \n for (const auto& window : weak_windows) {\n- if (window)\n+ if (window && !window->IsClosed())\n window->CloseImmediately... | 2022-02-23T15:27:54 |
facebook/react | 5c9614785c35a60d5f25ce8dcb514e91608ede8e | c72452895a43d3317590b03092ebf7c2fc9e3d34 | [codegen] Create new temporary map when entering reactive scopes
During codegen, when we now cache and restore the temporary values map as we
enter and exit the scope. This ensures that any temporaries within the reactive
scope are only visible within that scope, and not to subsequent code. In a
subsequent PR this... | [
{
"path": "compiler/forget/src/ReactiveScopes/CodegenReactiveFunction.ts",
"patch": "@@ -117,7 +117,9 @@ function codegenBlock(cx: Context, block: ReactiveBlock): t.BlockStatement {\n break;\n }\n case \"scope\": {\n+ const temp = new Map(cx.temp);\n codegenReactiveScope(c... | 2023-04-20T23:48:26 |
ggml-org/llama.cpp | 69e0ecef06bfef7d35c91f791b78651d4a3422e3 | 062cca58fcb085b9cd2abdf489588f76dcb702dd | webui: Fix editing assistant message without branching (#20944)
* fix: Editing assistant response without branching
* chore: update webui build output | [
{
"path": "tools/server/webui/src/lib/stores/chat.svelte.ts",
"patch": "@@ -1207,7 +1207,6 @@ class ChatStore {\n \t\t\t\tawait conversationsStore.updateCurrentNode(newMessage.id);\n \t\t\t} else {\n \t\t\t\tawait DatabaseService.updateMessage(msg.id, { content: newContent });\n-\t\t\t\tawait conversationsS... | 2026-03-25T10:47:33 |
ollama/ollama | ce929984a33230269905e0e3cfa335cb8d6ba781 | 4b34930a31ceb9cc10d95b8bcd60c319f47d8043 | server/internal/client/ollama: fix file descriptor management in Pull (#9931)
Close chunked writers as soon as downloads complete, rather than
deferring closure until Pull exits. This prevents exhausting file
descriptors when pulling many layers.
Instead of unbounded defers, use a WaitGroup and background goroutine
t... | [
{
"path": "server/internal/client/ollama/registry.go",
"patch": "@@ -486,44 +486,43 @@ func (r *Registry) Pull(ctx context.Context, name string) error {\n \t\texpected += l.Size\n \t}\n \n-\tvar total atomic.Int64\n+\tvar received atomic.Int64\n \tvar g errgroup.Group\n \tg.SetLimit(r.maxStreams())\n \tfor ... | 2025-03-21T23:16:38 |
huggingface/transformers | 33a8e68d4955c13cf6d6c629e76f12fc81974fc5 | 3b4a064d4ec700a3d6432e783a61e3450d5d8acc | Add GlmMoeDsa (#43858)
* draft
* for review only
* update ignore layers
* add config
* update
* rename
* fallback
* update
* 1
* update
* fix attention and date
* remove pretraining_tp and improve tests
* style
* remove pretraining_tp
* fix config
* fix compile
* remove wrong integration test
* fix
*... | [
{
"path": "docs/source/en/_toctree.yml",
"patch": "@@ -565,6 +565,8 @@\n title: GLM-4.7-Flash\n - local: model_doc/glm_image\n title: GLM-Image\n+ - local: model_doc/glm_moe_dsa\n+ title: GlmMoeDsa\n - local: model_doc/openai-gpt\n title: GPT\n - local: ... | 2026-02-09T12:06:23 |
golang/go | a86ea80197ed3bf0f276638a9ce079cbd2071d83 | a3a31ec8ed4aff33d9f5ef8358f9f74efb249f74 | cmd/go/internal/tool: set Internal.ExeName on tool's package
While the cached name of an executable is set based on the base name of
the package path, the executable produced as the output of link doesn't
have ExeName set on it and is just called a.out (with a .exe suffix on
Windows). Set ExeName so that the first tim... | [
{
"path": "src/cmd/go/internal/tool/tool.go",
"patch": "@@ -287,6 +287,7 @@ func buildAndRunModtool(ctx context.Context, tool string, args []string) {\n \tpkgOpts := load.PackageOpts{MainOnly: true}\n \tp := load.PackagesAndErrors(ctx, pkgOpts, []string{tool})[0]\n \tp.Internal.OmitDebug = true\n+\tp.Intern... | 2024-11-21T18:16:43 |
ggml-org/llama.cpp | 062cca58fcb085b9cd2abdf489588f76dcb702dd | 406f4e3f61f95a7ed95db81d1b5be0ec7e9bfb70 | Add SLEEPING status to the WebUI model selector (#20949)
* webui: handle sleeping model status, fix favourite -> favorite
* Update tools/server/webui/src/lib/components/app/models/ModelsSelectorOption.svelte
Co-authored-by: Aleksander Grygier <aleksander.grygier@gmail.com>
* Update tools/server/webui/src/lib/compon... | [
{
"path": "tools/server/webui/src/lib/components/app/models/ModelsSelector.svelte",
"patch": "@@ -77,7 +77,7 @@\n \tlet filteredOptions = $derived(filterModelOptions(options, searchTerm));\n \n \tlet groupedFilteredOptions = $derived(\n-\t\tgroupModelOptions(filteredOptions, modelsStore.favouriteModelIds, (... | 2026-03-25T10:02:32 |
electron/electron | 08e26175fdc8b7fb033e55c408181f2c0fa2940e | 41b2945ced7f93568fb3771ab6989d490214ab91 | fix: broken OSR transparent option (#32885) | [
{
"path": "shell/browser/api/electron_api_web_contents.cc",
"patch": "@@ -767,8 +767,12 @@ WebContents::WebContents(v8::Isolate* isolate,\n #if BUILDFLAG(ENABLE_OSR)\n }\n } else if (IsOffScreen()) {\n- bool transparent = false;\n- options.Get(options::kTransparent, &transparent);\n+ // webPr... | 2022-02-23T10:33:42 |
facebook/react | c72452895a43d3317590b03092ebf7c2fc9e3d34 | 8bab0bca3c1668afadd7c7e7261f934ee0b7ee18 | [codegen] validate that temporary values are set
In codegen when we lower an operand we check to see if we have an already
lowered value for it (stored in `cx.temp`). Currently we silently handle missing
values by emitting a raw identifier, but this is really an error. This PR adds
validation, which uncovered a fe... | [
{
"path": "compiler/forget/src/HIR/PrintHIR.ts",
"patch": "@@ -293,7 +293,11 @@ export function printInstructionValue(instrValue: ReactiveValue): string {\n }\n case \"JSXText\":\n case \"Primitive\": {\n- value = JSON.stringify(instrValue.value);\n+ if (instrValue.value === undefined)... | 2023-04-20T23:48:22 |
ollama/ollama | c794fef2f27d141393064665d2774b341d091393 | 00ebda8cc4f2031096973b26d6b0de7425a9ca82 | server/internal/client/ollama: persist through chunk download errors (#9923) | [
{
"path": "server/internal/client/ollama/registry.go",
"patch": "@@ -59,6 +59,11 @@ var (\n \t// ErrCached is passed to [Trace.PushUpdate] when a layer already\n \t// exists. It is a non-fatal error and is never returned by [Registry.Push].\n \tErrCached = errors.New(\"cached\")\n+\n+\t// ErrIncomplete is r... | 2025-03-21T20:03:43 |
golang/go | c37d9a0f9313ff2936b1b0b82476416e6cb03c62 | f5973c2183d64f93816c0a28189cdd3f263036fb | cmd/link: remove debugging dreg
I left this behind accidentally.
Change-Id: I70f97b8214775e89c612890aead26431c9a443a4
Reviewed-on: https://go-review.googlesource.com/c/go/+/630575
Auto-Submit: Russ Cox <rsc@golang.org>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-... | [
{
"path": "src/cmd/link/internal/ld/data.go",
"patch": "@@ -1089,7 +1089,6 @@ func writeBlock(ctxt *Link, out *OutBuf, ldr *loader.Loader, syms []loader.Sym,\n \t\t}\n \t\tif val < addr {\n \t\t\tldr.Errorf(s, \"phase error: addr=%#x but val=%#x sym=%s type=%v sect=%v sect.addr=%#x prev=%s\", addr, val, ldr... | 2024-11-21T14:10:36 |
ggml-org/llama.cpp | 406f4e3f61f95a7ed95db81d1b5be0ec7e9bfb70 | 53dc8b59bf51eafe5d5d0be7b8d3a6039abf6862 | android : fix-pointer-dangling (#20974) | [
{
"path": "examples/llama.android/lib/src/main/cpp/ai_chat.cpp",
"patch": "@@ -365,13 +365,13 @@ Java_com_arm_aichat_internal_InferenceEngineImpl_processSystemPrompt(\n const auto *system_prompt = env->GetStringUTFChars(jsystem_prompt, nullptr);\n LOGd(\"%s: System prompt received: \\n%s\", __func__... | 2026-03-25T09:51:26 |
facebook/react | cc93a8533264618b7a8e3d6fb56df917214c19d7 | fdad813ac765e901e2957b8d36fba87e5504e5f4 | [Fiber] InvokeGuardedCallback without metaprogramming (#26569)
InvokeGuardedCallback is now implemented with the browser fork done at
error-time rather than module-load-time. Originally it also tried to
freeze the window/document references to avoid mismatches in prototype
chains when testing React in different doc... | [
{
"path": "packages/shared/invokeGuardedCallbackImpl.js",
"patch": "@@ -7,80 +7,46 @@\n * @flow\n */\n \n-// $FlowFixMe[missing-this-annot]\n-function invokeGuardedCallbackProd<Args: Array<mixed>, Context>(\n- name: string | null,\n- func: (...Args) => mixed,\n- context: Context,\n-): void {\n- // $Fl... | 2023-04-20T22:08:51 |
electron/electron | 41b2945ced7f93568fb3771ab6989d490214ab91 | bcf060fab67fe1912d03ad253de98b3091b2a61b | feat: add ability to configure if window should close when opener closes (#31314)
* feat: Added ability to configure if window should close when opener closes
* fix: check if embedder is destroyed
* fix: correctly take over closeWithOpener property
* chore: Added documentation
* Update docs/api/window-open... | [
{
"path": "docs/api/window-open.md",
"patch": "@@ -73,6 +73,11 @@ creating the window. Note that this is more powerful than passing options\n through the feature string, as the renderer has more limited privileges in\n deciding security preferences than the main process.\n \n+In addition to passing in `acti... | 2022-02-23T07:59:50 |
ollama/ollama | 2d6eac9084a29060ccff69014e28e206a3a7a663 | 3ed7ad3ab32b458aa2fdb8d0144c546efdb26a72 | kvcache: Optimize sliding window attention
Currently sliding window attention allocates and uses the full
context size and just masks out any tokens that are outside of the
window. However, we really only need (roughly) the sliding window
size.
At large context sizes this improves two things:
- Memory allocated - si... | [
{
"path": "kvcache/causal.go",
"patch": "@@ -118,7 +118,12 @@ func (c *Causal) Init(backend ml.Backend, dtype ml.DType, maxSequences, capacity\n \t\tc.config.MaskDType = ml.DTypeF32\n \t}\n \n-\tcacheSize := maxSequences * capacity\n+\tvar cacheSize int\n+\tif c.windowSize == math.MaxInt32 || capacity < int... | 2025-03-18T20:13:32 |
golang/go | f5973c2183d64f93816c0a28189cdd3f263036fb | 4671276c5d5f2e51dd43e856267eac44cff18652 | cmd/go: fix -changed don`t print when GOFIPS140 is non-default
See https://github.com/golang/go/issues/70200#issuecomment-2468562595,
GOFIPS140 value when building the toolchain (off when not set)
is the default value for GOFIPS140,
it is buildcfg.defaultGOFIPS140,
export as buildcfg.DefaultGOFIPS140 that can be used ... | [
{
"path": "src/cmd/dist/buildruntime.go",
"patch": "@@ -65,7 +65,7 @@ func mkbuildcfg(file string) {\n \tfmt.Fprintf(&buf, \"const version = `%s`\\n\", findgoversion())\n \tfmt.Fprintf(&buf, \"const defaultGOOS = runtime.GOOS\\n\")\n \tfmt.Fprintf(&buf, \"const defaultGOARCH = runtime.GOARCH\\n\")\n-\tfmt.F... | 2024-11-20T16:01:36 |
huggingface/transformers | fc9137225880a9d03f130634c20f9dbe36a7b8bf | 161c43032b818fcf4b0a42f42f93e0a408d98d2c | Adding Support for Qwen3.5 (#43830)
* rebase main
* remove redundant init
* fix
* remove qwen3vlmoe mapping since resolved
* add auto image processor
* fix
* fix
* update qwen3_next_style rmsnorm
* update text config check
* simplify vision model
* simplify vision config
* inherit pretrainedmodel and qwen3n... | [
{
"path": "docs/source/en/_toctree.yml",
"patch": "@@ -721,6 +721,10 @@\n title: Qwen2MoE\n - local: model_doc/qwen3\n title: Qwen3\n+ - local: model_doc/qwen3_5\n+ title: Qwen3.5\n+ - local: model_doc/qwen3_5_moe\n+ title: Qwen3.5 Moe\n - local: model_doc... | 2026-02-09T11:21:25 |
ggml-org/llama.cpp | 53dc8b59bf51eafe5d5d0be7b8d3a6039abf6862 | 403c9c9cef26b875691f1ec52a7aa73b464dfba8 | sycl : fix wrong variable check by assert (#20903)
* fix wrong variable check by assert
* use GGML api | [
{
"path": "ggml/src/ggml-sycl/add-id.cpp",
"patch": "@@ -56,7 +56,7 @@ void ggml_sycl_add_id(ggml_backend_sycl_context& ctx, ggml_tensor* dst) {\n float* dst_d = (float*)dst->data;\n \n const unsigned int max_work_group_size = ggml_sycl_info().max_work_group_sizes[ctx.device];\n- assert(work_group_size... | 2026-03-25T09:48:37 |
facebook/react | fdad813ac765e901e2957b8d36fba87e5504e5f4 | e5708b3ea9190c1285c9081ff338e46be9ff39bc | [Float][Fiber] Enable Float methods to be called outside of render (#26557)
Stacked on #26570
Previously we restricted Float methods to only being callable while
rendering. This allowed us to make associations between calls and their
position in the DOM tree, for instance hoisting preinitialized styles
into a S... | [
{
"path": "packages/react-art/src/ReactFiberConfigART.js",
"patch": "@@ -479,11 +479,3 @@ export function suspendInstance(type, props) {}\n export function waitForCommitToBeReady() {\n return null;\n }\n-// eslint-disable-next-line no-undef\n-export function prepareRendererToRender(container: Container): ... | 2023-04-20T21:40:25 |
ollama/ollama | 6d1103048eac63f27148d6d8fe47c98cbb6f184f | 0ff28758b3a5e9dc0149a93d87677dd0585590c1 | fix: show correct bool value for kv in verbose show information (#9928) | [
{
"path": "cmd/cmd.go",
"patch": "@@ -703,6 +703,8 @@ func showInfo(resp *api.ShowResponse, verbose bool, w io.Writer) error {\n \t\t\tfor _, k := range keys {\n \t\t\t\tvar v string\n \t\t\t\tswitch vData := resp.ModelInfo[k].(type) {\n+\t\t\t\tcase bool:\n+\t\t\t\t\tv = fmt.Sprintf(\"%t\", vData)\n \t\t\t... | 2025-03-21T18:13:54 |
huggingface/transformers | 161c43032b818fcf4b0a42f42f93e0a408d98d2c | 41833e3e8c3b8c842751561aab62e308f625fd6e | fix gptoss crash in tp (#43853)
Signed-off-by: Wang, Yi <yi.a.wang@intel.com>
Co-authored-by: Mohamed Mekkouri <93391238+MekkCyber@users.noreply.github.com>
Co-authored-by: Marc Sun <57196510+SunMarc@users.noreply.github.com> | [
{
"path": "src/transformers/integrations/mxfp4.py",
"patch": "@@ -443,7 +443,7 @@ def mlp_forward(self, hidden_states):\n with on_device(router_logits.device):\n routing_data, gather_idx, scatter_idx = routing(router_logits, self.router.top_k)\n \n- routed_out = self.experts(hidden_states, ro... | 2026-02-09T10:42:33 |
electron/electron | e9b9835feb32189164de0a8bd67c85de5cee1895 | fa7461685dfec20208b21b91accad641ec8c577d | chore: fix out-of-date chromium patch (#33030) | [
{
"path": "patches/chromium/fix_don_t_restore_maximized_windows_when_calling_showinactive.patch",
"patch": "@@ -7,10 +7,10 @@ This is a backport from Chromium of\n https://chromium-review.googlesource.com/c/chromium/src/+/3371573.\n \n diff --git a/ui/views/win/hwnd_message_handler.cc b/ui/views/win/hwnd_me... | 2022-02-22T07:24:11 |
golang/go | 4671276c5d5f2e51dd43e856267eac44cff18652 | 21b5f2637a37fe2fece7e6bf3d5f97a9520693ba | crypto/tls: FIPS 140-3 mode
Consolidates handling of FIPS 140-3 considerations for the tls package.
Considerations specific to certificates are now handled in tls instead
of x509 to limit the area-of-effect of FIPS as much as possible.
Boringcrypto specific prefixes are renamed as appropriate.
For #69536
Co-authored... | [
{
"path": "src/crypto/internal/boring/fipstls/stub.s",
"patch": "@@ -1,12 +0,0 @@\n-// Copyright 2017 The Go Authors. All rights reserved.\n-// Use of this source code is governed by a BSD-style\n-// license that can be found in the LICENSE file.\n-\n-//go:build boringcrypto\n-\n-// runtime_arg0 is declared... | 2024-11-18T21:18:56 |
ggml-org/llama.cpp | 9f102a1407ed5d73b8c954f32edab50f8dfa3f58 | 3fc6f1aed172602790e9088b57786109438c2466 | models : move the token embedding norms to the first layer (#20943)
* models : move the token embedding norms to the first layer
* cont : fix LLM_TENSOR_CONV1D + fix il indexing | [
{
"path": "src/llama-arch.cpp",
"patch": "@@ -2564,7 +2564,7 @@ static const std::map<llm_tensor, llm_tensor_info> LLM_TENSOR_INFOS = {\n {LLM_TENSOR_TOKEN_EMBD, {LLM_TENSOR_LAYER_INPUT, GGML_OP_GET_ROWS}},\n {LLM_TENSOR_POS_EMBD, {LLM_TENSOR_LAYER_INPUT, GGML_OP_GE... | 2026-03-24T15:00:30 |
huggingface/transformers | 41833e3e8c3b8c842751561aab62e308f625fd6e | ab830e2916e4c1bca0887fef8186d0f12a9239de | Add id and resume support to SwanLab integration (#43719)
Fixes #43698
- Add SWANLAB_RUN_ID env var to pass experiment ID for resumption
- Add SWANLAB_RESUME env var to control resume mode
- Default to resume='allow' when resume_from_checkpoint is set"
Co-authored-by: Marc Sun <57196510+SunMarc@users.noreply.github.c... | [
{
"path": "src/transformers/integrations/integration_utils.py",
"patch": "@@ -2236,6 +2236,12 @@ def setup(self, args, state, model, **kwargs):\n - **SWANLAB_API_HOST** (`str`, *optional*, defaults to `None`):\n API address for the SwanLab cloud environment for private version (its free)... | 2026-02-09T10:41:03 |
ollama/ollama | d3e9ca3eda5585204358f4ba23b3c7cb23bdb3e2 | 0fbfcf3c9c7bfdbf4616238595eafd7eca2a916c | kvcache: Account for source tensors in defrag operation count
Defragging the KV cache can generate a lot of operations, so we
need to be careful that we don't overflow the number that the graph
can support. We currently account for all of the nodes that we add
to the graph for each move but we also need to include the... | [
{
"path": "kvcache/causal.go",
"patch": "@@ -321,7 +321,8 @@ func (c *Causal) defrag() {\n \tctx := c.backend.NewContext()\n \n \t// For every move, 6 tensors are required per layer (2 views and a\n-\t// copy for each of k and v).\n+\t// copy for each of k and v). We also need to refer to the original\n+\t/... | 2025-03-20T21:27:17 |
facebook/react | e5708b3ea9190c1285c9081ff338e46be9ff39bc | d73d7d59086218b0fa42d0a79c32a0365952650b | [Tests][Fizz] Better HTML parsing behavior for Fizz tests (#26570)
In anticipation of making Fiber use the document global for dispatching
Float methods that arrive from Flight I needed to update some tests that
commonly recreated the JSDOM instance after importing react.
This change updates a few tests to only c... | [
{
"path": "packages/react-dom/src/__tests__/ReactDOMFizzServer-test.js",
"patch": "@@ -10,7 +10,7 @@\n \n 'use strict';\n import {\n- replaceScriptsAndMove,\n+ insertNodesAndExecuteScripts,\n mergeOptions,\n stripExternalRuntimeInNodes,\n withLoadingReadyState,\n@@ -29,8 +29,6 @@ let useSyncExternal... | 2023-04-20T21:27:02 |
ggml-org/llama.cpp | 42ebce3bebeab64fbe71b667d1bafd9960e83cbf | a94fdb090a7fb9192749386f846176a9d8c97681 | common : fix get_gguf_split_info (#20946)
Signed-off-by: Adrien Gallouët <angt@huggingface.co> | [
{
"path": "common/download.cpp",
"patch": "@@ -460,9 +460,9 @@ static gguf_split_info get_gguf_split_info(const std::string & path) {\n int count = 1;\n \n if (std::regex_match(prefix, m, re_split)) {\n- prefix = m[1].str();\n index = std::stoi(m[2].str());\n count = std::stoi... | 2026-03-24T12:33:14 |
electron/electron | 1e8da899a372c40eab6743fbff5ce846736d8927 | 069cde09fbae94e18bb1e31ed6235c0761c08315 | chore: fix typos (#32985) | [
{
"path": "lib/browser/api/net.ts",
"patch": "@@ -198,7 +198,7 @@ class ChunkedBodyStream extends Writable {\n this._downstream = pipe;\n if (this._pendingChunk) {\n const doneWriting = (maybeError: Error | void) => {\n- // If the underlying request has been aborted, we honeslty don't c... | 2022-02-21T09:27:45 |
golang/go | 21b5f2637a37fe2fece7e6bf3d5f97a9520693ba | 844426104a901b68f3a8c7a7ca96d910b03fb054 | runtime: utilize EVFILT_USER more effectively
Re-work kqueue_event wakeup logic to use one-shot events. In an
event of waking up a wrong thread, simply re-post the event.
This saves close to 1 system call per wakeup on average, since
chances of non-blocking poller picking it up is pretty low.
Change-Id: I202d0d57a31... | [
{
"path": "src/runtime/netpoll_kqueue.go",
"patch": "@@ -129,10 +129,11 @@ retry:\n \t\tev := &events[i]\n \n \t\tif isWakeup(ev) {\n-\t\t\tif delay != 0 {\n+\t\t\tisBlocking := delay != 0\n+\t\t\tprocessWakeupEvent(kq, isBlocking)\n+\t\t\tif isBlocking {\n \t\t\t\t// netpollBreak could be picked up by a no... | 2024-11-20T19:27:40 |
huggingface/transformers | 2c3926ec921980ae4e9755cd2ffc3daad7a0514d | e0cb4d18cea52f1e0c573eb821f21f7ff33b922c | Fix incorrect timestamp calculation in Qwen3VL Processor (#43659)
Fixes #43519
The Qwen3VL processor was incorrectly using 'merge_size' (spatial merging)
instead of 'temporal_patch_size' (temporal grouping) when calculating
video frame timestamps.
When calculating timestamps:
- 'temporal_patch_size' groups consecuti... | [
{
"path": "src/transformers/models/qwen3_vl/modular_qwen3_vl.py",
"patch": "@@ -1509,7 +1509,7 @@ def __call__(\n curr_timestamp = self._calculate_timestamps(\n metadata.frames_indices,\n metadata.fps,\n- self.video_p... | 2026-02-09T09:54:09 |
facebook/react | 8bab0bca3c1668afadd7c7e7261f934ee0b7ee18 | c228d63f0375b0051b4ce18d2e748886ae292ba8 | [babel] Compile individual components
Updates the Babel plugin so that we can individually compile components and skip
over ones that have non-critical errors | [
{
"path": "compiler/forget/src/Babel/BabelPlugin.ts",
"patch": "@@ -54,48 +54,67 @@ export default function ReactForgetBabelPlugin(\n fn: BabelCore.NodePath<t.FunctionDeclaration>,\n pass: BabelPluginPass\n ): void {\n- const compiled = compile(fn, pass.opts.environment);\n-\n- if (pass.opts... | 2023-04-20T20:21:40 |
ggml-org/llama.cpp | a94fdb090a7fb9192749386f846176a9d8c97681 | c9dc43333f7b56fb7b1f306771f4a477602f6d81 | WebUI: fix edit msg form textarea height (#20830)
* autoresize textarea on mount
* allow textarea to grow to same height as rendered messages
* add UI build file | [
{
"path": "tools/server/webui/src/lib/components/app/chat/ChatForm/ChatFormTextarea.svelte",
"patch": "@@ -26,6 +26,7 @@\n \n \tonMount(() => {\n \t\tif (textareaElement) {\n+\t\t\tautoResizeTextarea(textareaElement);\n \t\t\ttextareaElement.focus();\n \t\t}\n \t});\n@@ -50,8 +51,9 @@\n \t<textarea\n \t\tbi... | 2026-03-24T12:17:45 |
ollama/ollama | 42a14f7f633110ab83343848865d4612cfefb398 | f8c3dbe5b5ee342d97e4c71d684b85b00273c33d | sample: add error handling for empty logits (#9740) | [
{
"path": "sample/samplers.go",
"patch": "@@ -26,6 +26,10 @@ type Sampler struct {\n }\n \n func (s *Sampler) Sample(logits []float32) (int32, error) {\n+\tif len(logits) == 0 {\n+\t\treturn -1, errors.New(\"sample: no logits provided to sample\")\n+\t}\n+\n \ttokens := make([]token, len(logits))\n \tfor i ... | 2025-03-20T18:11:18 |
electron/electron | 069cde09fbae94e18bb1e31ed6235c0761c08315 | bdad6335c4e18aecbb3c4f3176fc100e542f0477 | fix: don't restore maximized BrowserWindow when calling showInactive (#32870) | [
{
"path": "patches/chromium/.patches",
"patch": "@@ -111,3 +111,4 @@ fix_aspect_ratio_with_max_size.patch\n fix_dont_delete_SerialPortManager_on_main_thread.patch\n feat_add_data_transfer_to_requestsingleinstancelock.patch\n fix_crash_when_saving_edited_pdf_files.patch\n+fix_don_t_restore_maximized_windows_... | 2022-02-21T09:23:55 |
huggingface/transformers | 9e5c7d16aa13cf4c160076d23ec6d4d2ef5a8fd0 | 7769f660935b5d48b73bf6711d0a78b6f8f98739 | Fix slack-report workflow file (#43851)
fix
Co-authored-by: ydshieh <ydshieh@users.noreply.github.com> | [
{
"path": ".github/workflows/slack-report.yml",
"patch": "@@ -94,6 +94,7 @@ jobs:\n pip install huggingface_hub\n pip install slack_sdk\n pip show slack_sdk\n+ pip install .\n if [ \"$quantization_matrix\" != \"\" ]; then\n python utils/notificati... | 2026-02-09T05:41:29 |
facebook/react | c228d63f0375b0051b4ce18d2e748886ae292ba8 | 275c7557156778168c455d6086846cbde9cb7adf | [babel] Add panicOnBailout option
Adds an option to always throw errors regardless of severity (default, ie the
status quo), or when the flag is disabled, only critical errors will be thrown.
Any error that isn't considered a critical error (see
`CompilerError.isCritical()`) since it might indicate that the compil... | [
{
"path": "compiler/forget/src/Babel/BabelPlugin.ts",
"patch": "@@ -54,7 +54,6 @@ export default function ReactForgetBabelPlugin(\n fn: BabelCore.NodePath<t.FunctionDeclaration>,\n pass: BabelPluginPass\n ): void {\n- hasForgetCompiledCode = true;\n const compiled = compile(fn, pass.opts.en... | 2023-04-20T20:21:40 |
ggml-org/llama.cpp | 8c7957ca33a40cd928146fd3f33a98180e486004 | e852eb490136dfa335586b9dd6efb94b02c8fdd3 | common : add standard Hugging Face cache support (#20775)
* common : add standard Hugging Face cache support
- Use HF API to find all files
- Migrate all manifests to hugging face cache at startup
Signed-off-by: Adrien Gallouët <angt@huggingface.co>
* Check with the quant tag
Signed-off-by: Adrien Gallouët <angt@h... | [
{
"path": "common/CMakeLists.txt",
"patch": "@@ -63,6 +63,8 @@ add_library(${TARGET} STATIC\n debug.h\n download.cpp\n download.h\n+ hf-cache.cpp\n+ hf-cache.h\n http.h\n json-partial.cpp\n json-partial.h",
"additions": 2,
"deletions": 0,
"language": "Plain Text"
... | 2026-03-24T06:30:33 |
electron/electron | bdad6335c4e18aecbb3c4f3176fc100e542f0477 | 9d72c8b0ad2ee41541b01ce6594012b40d869f0b | fix: command string for windows protocol handler (#32953) | [
{
"path": "shell/browser/browser_win.cc",
"patch": "@@ -67,7 +67,7 @@ bool GetProtocolLaunchPath(gin::Arguments* args, std::wstring* exe) {\n std::vector<std::wstring> launch_args;\n if (args->GetNext(&launch_args) && !launch_args.empty())\n *exe = base::StringPrintf(L\"\\\"%ls\\\" \\\"%ls\\\" \\\"%... | 2022-02-21T07:43:27 |
huggingface/transformers | 7769f660935b5d48b73bf6711d0a78b6f8f98739 | dd360ad2364382e7ef3c19d1011cb0a4e9b418ff | Fix markdown documentation (#43076)
* Fix markdown documentation
Signed-off-by: Yuanyuan Chen <cyyever@outlook.com>
* Update docs/source/en/main_classes/video_processor.md
Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com>
---------
Signed-off-by: Yuanyuan Chen <cyyever@outlook.com>
Co-author... | [
{
"path": "docs/source/en/main_classes/video_processor.md",
"patch": "@@ -22,7 +22,7 @@ The video processor extends the functionality of image processors by allowing Vi\n \n When adding a new VLM or updating an existing one to enable distinct video preprocessing, saving and reloading the processor configura... | 2026-02-07T02:17:54 |
facebook/react | 275c7557156778168c455d6086846cbde9cb7adf | 3aa6196e306363ad962ac5a9014e366f99e7fa35 | Repro for jsx lowering issues
* JSX tag value temporaries getting promoted due to being sandwiched inside the
mutation of some other item
* Incorrect order-of-evaluation for jsx tag relative to props/children | [
{
"path": "compiler/forget/src/__tests__/fixtures/compiler/_bug.global-jsx-tag-lowered-between-mutations.expect.md",
"patch": "@@ -0,0 +1,53 @@\n+\n+## Input\n+\n+```javascript\n+function Component(props) {\n+ const maybeMutable = new MaybeMutable();\n+ // NOTE: this will produce invalid output.\n+ // Th... | 2023-04-20T16:29:37 |
electron/electron | faa392af28110a91b4cd018675331524d8e69048 | 683ff2ea02dd1e343b03f0c55cf8c30dbb8fa720 | docs: fix broken link to GNOME notifications spec (#32934) | [
{
"path": "docs/tutorial/notifications.md",
"patch": "@@ -146,7 +146,7 @@ desktop environment that follows [Desktop Notifications\n Specification][notification-spec], including Cinnamon, Enlightenment, Unity,\n GNOME, KDE.\n \n-[notification-spec]: https://developer.gnome.org/notification-spec/\n+[notificat... | 2022-02-21T01:23:17 |
golang/go | c1a0ee37644559f51aeb112dcf6ce3e4316a44df | f7b15264c8656cdda5db1364f30f813dabd743f4 | cmd/go: sort "no test files" test results into normal ordering
The code takes care to print test results during "go test ./..."
in the package order, delaying prints until it's that package's
turn, even when tests run in parallel. For some reason, the
prints about the test not running were not included in that,
making... | [
{
"path": "src/cmd/go/internal/test/test.go",
"patch": "@@ -1387,6 +1387,27 @@ func (r *runTestActor) Act(b *work.Builder, ctx context.Context, a *work.Action)\n \t\treturn nil\n \t}\n \n+\t// Stream test output (no buffering) when no package has\n+\t// been given on the command line (implicit current direc... | 2024-11-21T13:00:06 |
ggml-org/llama.cpp | e852eb490136dfa335586b9dd6efb94b02c8fdd3 | 312d870a8988f73f14a46075ad6a1ac1c861fafd | llama-fit: fix regex pattern for gate_up tensors (#20910)
* llama-fit: fix regex pattern for gate_up tensors
* Apply suggestions from code review
Co-authored-by: Johannes Gäßler <johannesg@5d6.de>
---------
Co-authored-by: Johannes Gäßler <johannesg@5d6.de> | [
{
"path": "src/llama.cpp",
"patch": "@@ -365,14 +365,14 @@ static void llama_params_fit_impl(\n case LAYER_FRACTION_ATTN: {\n static std::array<std::string, n_strings> patterns;\n if (patterns[il].empty()) {\n- patterns[il] = \"blk\\\\.\" + std:... | 2026-03-24T04:57:57 |
huggingface/transformers | f70ec1e7fede93598d9eb93d13f612c75d294d3a | 5ba28289d7f04f1b0d0c24bb596ab972b32f0d49 | Fix DeepSpeed model preparation logic in Trainer class (#43780)
Fix deepspeed model preparation logic in Trainer class
Co-authored-by: Marc Sun <57196510+SunMarc@users.noreply.github.com> | [
{
"path": "src/transformers/trainer.py",
"patch": "@@ -2297,6 +2297,8 @@ def _inner_training_loop(\n model, self.optimizer, self.lr_scheduler = self.accelerator.prepare(\n self.model, self.optimizer, self.lr_scheduler\n )\n+ else... | 2026-02-06T19:20:00 |
facebook/react | 7f8c501f682bd4abe24826a93538059d717ba39e | 22d5942675fcbd8b15b532284b49db4cb00d7144 | React DevTools 4.27.5 -> 4.27.6 (#26684)
Full list of changes:
* Use .slice() for all substring-ing
([sophiebits](https://github.com/sophiebits) in
[#26677](https://github.com/facebook/react/pull/26677))
* cleanup[devtools]: remove named hooks & profiler changed hook indices
feature flags ([hoxyq](https://github.... | [
{
"path": "packages/react-devtools-core/package.json",
"patch": "@@ -1,6 +1,6 @@\n {\n \"name\": \"react-devtools-core\",\n- \"version\": \"4.27.5\",\n+ \"version\": \"4.27.6\",\n \"description\": \"Use react-devtools outside of the browser\",\n \"license\": \"MIT\",\n \"main\": \"./dist/backend.j... | 2023-04-20T12:34:25 |
electron/electron | b1463d2da138dcbc20d87cf31b9807d697690592 | 36e730da9367b6e75e9fed882b9e8d652226919d | fix: stale renderer process on quit (#32888) | [
{
"path": "shell/browser/electron_browser_main_parts.cc",
"patch": "@@ -458,7 +458,8 @@ void ElectronBrowserMainParts::WillRunMainMessageLoop(\n std::unique_ptr<base::RunLoop>& run_loop) {\n js_env_->OnMessageLoopCreated();\n exit_code_ = content::RESULT_CODE_NORMAL_EXIT;\n- Browser::Get()->SetMain... | 2022-02-18T05:02:22 |
ggml-org/llama.cpp | 312d870a8988f73f14a46075ad6a1ac1c861fafd | 7cadbfce10fc16032cfb576ca4607cd2dd183bf1 | common : replace wrap_for_generation with a prefix convenience function and fix gpt-oss (#20912) | [
{
"path": "common/chat-auto-parser-generator.cpp",
"patch": "@@ -112,8 +112,7 @@ common_peg_arena autoparser::build_parser(const generation_params & inputs) cons\n } else {\n parser = content.build_parser(ctx);\n }\n- parser = wrap_for_generation_prompt(p, parser, inputs, ... | 2026-03-24T03:21:47 |
golang/go | f7b15264c8656cdda5db1364f30f813dabd743f4 | ce33585a1a06013f197ed95cc750e01bde5319c2 | cmd/dist: vet non-test packages in short mode
Pass all packages to go test, even if they don't have test files,
so that go test can still run vet.
I just got burned by a vet error in a package without a test
showing up when I added an (unrelated) test.
There are not enough packages without tests to be worth
the "savi... | [
{
"path": "src/cmd/dist/test.go",
"patch": "@@ -641,19 +641,17 @@ func (t *tester) registerTests() {\n \t\t// Use 'go list std cmd' to get a list of all Go packages\n \t\t// that running 'go test std cmd' could find problems in.\n \t\t// (In race test mode, also set -tags=race.)\n-\t\t//\n-\t\t// In long te... | 2024-11-20T13:01:57 |
ollama/ollama | df94175a0fb0356c9b9e9a62b73d908633c08810 | 61a882521657a27b5e1a0834eb6ab69617290ef2 | ggml: return error on failure to read tensor data (#9872)
When converting a ggml model if there is a failure to read tensor data a nil error value was being returned. It should be assigned to the actual error from reading. | [
{
"path": "ml/backend/ggml/ggml.go",
"patch": "@@ -330,7 +330,7 @@ func New(r *os.File, params ml.BackendParams) (ml.Backend, error) {\n \t\t}\n \t}\n \n-\tif g.Wait() != nil {\n+\tif err := g.Wait(); err != nil {\n \t\treturn nil, err\n \t}\n ",
"additions": 1,
"deletions": 1,
"language": "Go"
... | 2025-03-18T23:51:33 |
huggingface/transformers | 0c89522f2af2f85cf997193645a1e727d6b8c1d7 | e61eed44ba035f019ae381f1b7c0c06dbf78d132 | fix(flaky): enforce manual seed to reduce flakiness (#43794)
This change aims to reduce flakiness in CI tests. We identified two causes of nondeterministic behavior:
- Some tests were not using a fixed RNG seed, which reduced determinism.
- The cli tests were occasionally triggering I/O errors due to writes on a clo... | [
{
"path": "tests/causal_lm_tester.py",
"patch": "@@ -604,6 +604,8 @@ def test_flash_attn_2_equivalence(self):\n if not model_class._supports_flash_attn:\n self.skipTest(reason=\"Model does not support Flash Attention 2\")\n \n+ # Set seed for deterministic test - ensur... | 2026-02-06T16:30:28 |
ggml-org/llama.cpp | 7cadbfce10fc16032cfb576ca4607cd2dd183bf1 | 1fb2290a51f04cc79e7b162713e13ccf351f9de3 | hexagon: general DMA and Binary Op fixes for large strides (#20918)
* hex-dma: make chained dma the default to handle newer models
This also includes some new instrumentation that we can remove later.
* hexagon: add uint32 dump helper
* hexagon: use single-page VTCM allocation to avoid issues with large gather ops ... | [
{
"path": "ggml/src/ggml-hexagon/ggml-hexagon.cpp",
"patch": "@@ -461,7 +461,7 @@ static void repack_row_q4x4x2(uint8_t * y, const block_q4_0 * x, int64_t k) {\n d[7] = x[i * 8 + 7].d;\n }\n \n- if (opt_verbose > 1) {\n+ if (opt_verbose > 2) {\n for (int i = 0; i < nb; i++... | 2026-03-23T22:33:49 |
golang/go | ce33585a1a06013f197ed95cc750e01bde5319c2 | 02df9bbdbdb0a46b86b34d0cf236b599311c0ecb | cmd/dist: adjust FIPS 140 tests
An earlier CL moved the actual test from crypto/internal/fips/check
to crypto/internal/fipstest (now crypto/internal/fips140test),
so this cmd/dist check has been doing nothing for a little while.
Fix it to do what it intends.
Also run the actual crypto package tests in FIPS mode in lo... | [
{
"path": "src/cmd/dist/test.go",
"patch": "@@ -323,6 +323,7 @@ type goTest struct {\n \tbench bool // Run benchmarks (briefly), not tests.\n \trunTests string // Regexp of tests to run\n \tcpu string // If non-empty, -cpu flag\n+\tskip string // If non-empty, -skip... | 2024-11-20T14:23:40 |
ollama/ollama | 61a882521657a27b5e1a0834eb6ab69617290ef2 | 021dcf089d77292976ee7655eca424dd0b53b8f4 | convert: return name of unsupported architecture (#9862)
When a model's architecture cannot be converted return the name of the unsupported arch in the error message. | [
{
"path": "convert/convert.go",
"patch": "@@ -201,7 +201,7 @@ func ConvertModel(fsys fs.FS, ws io.WriteSeeker) error {\n \tcase \"CohereForCausalLM\":\n \t\tconv = &commandrModel{}\n \tdefault:\n-\t\treturn errors.New(\"unsupported architecture\")\n+\t\treturn fmt.Errorf(\"unsupported architecture %q\", p.A... | 2025-03-18T17:38:28 |
electron/electron | e08ced59790089b84cd7efa81a983155faa82c5f | 84cf685e761b2dd22f064e0dbf33b290bbec5fc0 | chore: update patch to fix CI (#32950) | [
{
"path": "patches/devtools_frontend/fix_expose_globals_to_allow_patching_devtools_dock.patch",
"patch": "@@ -36,10 +36,10 @@ index 89255a0927a647ca32d1a9508853425a3207b441..f0e1e32f80d79e400ad139818edce60a\n import '../../ui/legacy/components/source_frame/source_frame-meta.js';\n import '../../panels/con... | 2022-02-17T12:05:01 |
facebook/react | 767f52237cf7892ad07726f21e3e8bacfc8af839 | c826dc50de288758a0b783b2fd37b40a3b512fc4 | Use .slice() for all substring-ing (#26677)
- substr is Annex B
- substring silently flips its arguments if they're in the "wrong order", which is confusing
- slice is better than sliced bread (no pun intended) and also it works the same way on Arrays so there's less to remember
---
> I'd be down to just lint ... | [
{
"path": ".eslintrc.js",
"patch": "@@ -236,7 +236,14 @@ module.exports = {\n 'no-inner-declarations': [ERROR, 'functions'],\n 'no-multi-spaces': ERROR,\n 'no-restricted-globals': [ERROR].concat(restrictedGlobals),\n- 'no-restricted-syntax': [ERROR, 'WithStatement'],\n+ 'no-restricted-synt... | 2023-04-19T21:26:01 |
huggingface/transformers | f8f2834e1a82970f93e06f9bb076a361ca2d70c0 | 95007e597bf7bc73774a2877723bc182f0e0a43b | [core] Faster and thread-safe `check_model_inputs` implementation (#43765)
* poc
* cleaner
* style
* forgot name
* fix all
* more fixes
* oupsi
* fix
* use only 1 contextvar
* lazy hook installation
* capture with lists instead of tuples
* more bad-behaved model fixes
* style
* move to dedicated file
* d... | [
{
"path": "src/transformers/modeling_utils.py",
"patch": "@@ -118,14 +118,15 @@\n is_torch_xpu_available,\n logging,\n )\n-from .utils.generic import _CAN_RECORD_REGISTRY, GeneralInterface, OutputRecorder, is_flash_attention_requested\n+from .utils.generic import GeneralInterface, is_flash_attention... | 2026-02-06T16:08:46 |
ggml-org/llama.cpp | 1772701f99dd3fc13f5783b282c2361eda8ca47c | 39bf0d3c6a95803e0f41aaba069ffbee26721042 | opencl: add q6_K gemm and gemv kernels for Adreno (#20089)
* opencl: add q6_K noshuffle kernels, initial q6_K gemv, some host code
* opencl: add q6_K transpose
* opencl: fix cvt kernel name
* opencl: add call to q6_K gemv
* opencl: fix q6_K scale transpose
* opencl: fix loading for gemv q6_K, refactor
* opencl: ... | [
{
"path": "ggml/src/ggml-opencl/CMakeLists.txt",
"patch": "@@ -114,6 +114,8 @@ set(GGML_OPENCL_KERNELS\n gemv_noshuffle_q4_1_f32\n gemm_noshuffle_q4_1_f32\n gemv_noshuffle_general_q8_0_f32\n+ gemv_noshuffle_q6_k_f32\n+ gemm_noshuffle_q6_k_f32\n mul\n neg\n norm",
"additions... | 2026-03-23T19:44:18 |
ollama/ollama | bf24498b1ea2728ef735030f186d1c970c38eaa6 | 95e271d98f0b5717bde2b7e5c4c7639c19fd3763 | ollamarunner: Check for minBatch of context space when shifting
Models can specify that a group of inputs need to be handled a single
batch. However, context shifting didn't respect this and could trigger
a break anyways. In this case, we should instead trigger a context
shift earlier so that it occurs before the grou... | [
{
"path": "runner/ollamarunner/runner.go",
"patch": "@@ -115,6 +115,9 @@ func (s *Server) NewSequence(prompt string, images []llm.ImageData, params NewSe\n \t\tparams.numKeep = int32(len(inputs))\n \t}\n \n+\t// TODO(jessegross): We should ensure that we always leave minBatch of context space to shift,\n+\t... | 2025-03-17T19:03:06 |
golang/go | 6daf022e0f4f8ad5ceeaebb06de287c84113c311 | a5c1b1c43389df3cb108e333a4198c81f15b1140 | cmd/go remove base.ShortPathConservative
This CL rolls forward CL 630276, fixing the issues with the longtest builders that required the revert in CL 630317.
The change this CL makes compared to CL 630276 is adding the
shortPathErrorList function to rewrite the paths in the modfile.Errors
in a modfile.ErrorList using... | [
{
"path": "src/cmd/go/internal/base/path.go",
"patch": "@@ -6,13 +6,12 @@ package base\n \n import (\n \t\"errors\"\n+\t\"io/fs\"\n \t\"os\"\n \t\"path/filepath\"\n \t\"runtime\"\n \t\"strings\"\n \t\"sync\"\n-\n-\t\"cmd/go/internal/str\"\n )\n \n // UncachedCwd returns the current working directory.\n@@ -3... | 2024-11-21T14:55:32 |
electron/electron | 84cf685e761b2dd22f064e0dbf33b290bbec5fc0 | fe43296f7fd11171ebd73c0e6d248b9c6da5279d | fix: webContents.openDevTools({mode}) not working (#32829) | [
{
"path": "patches/config.json",
"patch": "@@ -3,6 +3,8 @@\n \n \"src/electron/patches/boringssl\": \"src/third_party/boringssl/src\",\n \n+ \"src/electron/patches/devtools_frontend\": \"src/third_party/devtools-frontend/src\",\n+\n \"src/electron/patches/webrtc\": \"src/third_party/webrtc\",\n \n \"... | 2022-02-17T05:59:13 |
facebook/react | 3aa6196e306363ad962ac5a9014e366f99e7fa35 | a0777212df7009cc26e47ffd46444276d0be923d | Starting point for a new, parallel test runner to replace our hacked Jest
snapshotting
As demo'd on our sync. This is meant as a replacement for `yarn test` just for
fixtures. On my machine, a test run from a steady state of Jest watch mode takes
6 seconds. With this script, it takes ~800ms.
Workflow: make edits... | [
{
"path": "compiler/forget/packages/snap/package.json",
"patch": "@@ -0,0 +1,39 @@\n+{\n+ \"name\": \"snap\",\n+ \"version\": \"0.0.1\",\n+ \"public\": false,\n+ \"description\": \"Snapshot testing CLI tool\",\n+ \"main\": \"dist/index.js\",\n+ \"license\": \"MIT\",\n+ \"files\": [\n+ \"src\"\n+ ... | 2023-04-19T21:07:06 |
ollama/ollama | 2c8b4846437747bd23e7a176f83011e39ec2128b | 8294676150dde3dd3d316644b86e42b07204ff9c | fix: correctly save in interactive mode (#9788)
This fixes the case where a FROM line in previous modelfile points to a
file which may/may not be present in a different ollama instance. We
shouldn't be relying on the filename though and instead just check if
the FROM line was instead a valid model name and point to th... | [
{
"path": "cmd/cmd_test.go",
"patch": "@@ -757,3 +757,132 @@ func TestCreateHandler(t *testing.T) {\n \t\t})\n \t}\n }\n+\n+func TestNewCreateRequest(t *testing.T) {\n+\ttests := []struct {\n+\t\tname string\n+\t\tfrom string\n+\t\topts runOptions\n+\t\texpected *api.CreateRequest\n+\t}{\n+\t\t{... | 2025-03-15T19:09:02 |
golang/go | 0a0a7a564271ab8acfe6210a6e1ca19e712e0d1f | 3ff868f2f50ed5ec44f77bf9f27e42e51e2aae4a | cmd/compile: fix rewrite rules for multiply/add
x - (y - c) == (x - y) + c, not (x - y) - c. Oops.
Fixes #70481
Change-Id: I0e54d8e65dd9843c6b92c543ac69d69ee21f617c
Reviewed-on: https://go-review.googlesource.com/c/go/+/630397
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
... | [
{
"path": "src/cmd/compile/internal/ssa/_gen/ARM64.rules",
"patch": "@@ -1150,8 +1150,8 @@\n // madd/msub can't take constant arguments, so do a bit of reordering if a non-constant is available.\n (ADD a p:(ADDconst [c] m:((MUL|MULW|MNEG|MNEGW) _ _))) && p.Uses==1 && m.Uses==1 => (ADDconst [c] (ADD <v.Type>... | 2024-11-21T01:48:30 |
electron/electron | fe43296f7fd11171ebd73c0e6d248b9c6da5279d | 34129b83a471cbe2aa4de153c1af084a1cd8be83 | chore: backport EPROTOTYPE fixes from libuv (#32856)
This commit backports three commits from libuv's 1.x branch to fix
issues with CPU going to 100% on macOS when EPROTOTYPE is returned.
See: https://github.com/libuv/libuv/commit/abb109f30f7c0b3615b75156376d2e886c365df8
See: https://github.com/libuv/libuv/commit... | [
{
"path": "patches/node/.patches",
"patch": "@@ -28,3 +28,6 @@ fix_crash_caused_by_gethostnamew_on_windows_7.patch\n fix_suppress_clang_-wdeprecated-declarations_in_libuv.patch\n fix_don_t_create_console_window_when_creating_process.patch\n fix_serdes_test.patch\n+darwin_remove_eprototype_error_workaround_3... | 2022-02-17T05:58:21 |
ggml-org/llama.cpp | f93c09e2673090b59802f2552848a1cac70598b0 | 841bc203e2fdd3bcc032277766984bd5a35d7c1d | memory : fix seq_id bounds in llama_memory_recurrent::state_read_meta() (#20887) | [
{
"path": "src/llama-memory-recurrent.cpp",
"patch": "@@ -928,11 +928,8 @@ bool llama_memory_recurrent::state_read_meta(llama_io_read_i & io, uint32_t cell\n llama_seq_id seq_id;\n io.read_to(&seq_id, sizeof(seq_id));\n \n- // TODO: llama_memory_recurrent shoul... | 2026-03-23T12:08:46 |
facebook/react | c826dc50de288758a0b783b2fd37b40a3b512fc4 | cd2b79dedd6d81abb0b01d38396afa083feaf9e9 | Add (Client) Functions as Form Actions (#26674)
This lets you pass a function to `<form action={...}>` or `<button
formAction={...}>` or `<input type="submit formAction={...}>`. This will
behave basically like a `javascript:` URL except not quite implemented
that way. This is a convenience for the `onSubmit={e => {... | [
{
"path": "fixtures/flight/src/Button.js",
"patch": "@@ -6,20 +6,22 @@ export default function Button({action, children}) {\n const [isPending, setIsPending] = React.useState(false);\n \n return (\n- <button\n- disabled={isPending}\n- onClick={async () => {\n- setIsPending(true);\n- ... | 2023-04-19T20:31:08 |
ollama/ollama | 7bf793a6007ca11fae0180ea6f2ebd7258428bd4 | 282bfaaa957fff8402eb7a4b1657f57c49939604 | gemma3: Allow multiple image in a single input
Previously processing multiple images in a batch would trigger
segfaults so sending images together was disabled as a way to
mitigate this. The trigger was processing one image on the CPU
and one on the GPU.
This can no longer happen:
- The vision encoder is now on the ... | [
{
"path": "server/prompt.go",
"patch": "@@ -26,7 +26,6 @@ func chatPrompt(ctx context.Context, m *Model, tokenize tokenizeFunc, opts *api.\n \tvar system []api.Message\n \n \tisMllama := checkMllamaModelFamily(m)\n-\tisGemma3 := checkGemma3ModelFamily(m)\n \n \tvar imageNumTokens int\n \t// TODO: Ideally we... | 2025-03-12T23:59:23 |
golang/go | 3ff868f2f50ed5ec44f77bf9f27e42e51e2aae4a | 28f4e14ebe281d8e46cba430bfd123ce21fcd0cc | cmd/go: cache executables built for go run
This change implements executable caching. It always caches the outputs of
link steps used by go run. To do so we need to make a few changes:
The first is that we want to cache binaries in a slightly different
location than we cache other outputs. The reason for doing so is ... | [
{
"path": "src/cmd/go/internal/base/base.go",
"patch": "@@ -17,6 +17,7 @@ import (\n \t\"slices\"\n \t\"strings\"\n \t\"sync\"\n+\t\"time\"\n \n \t\"cmd/go/internal/cfg\"\n \t\"cmd/go/internal/str\"\n@@ -206,18 +207,34 @@ func Run(cmdargs ...any) {\n \t}\n }\n \n-// RunStdin is like run but connects Stdin.\... | 2024-09-13T15:06:06 |
ggml-org/llama.cpp | c44a932cf495dc1cf73020f62acce6233a0a95f3 | 177c75852a6a6fd6810ce23f50c77e4a60ea6828 | webui: fix --webui-config-file settings not applied on load (#20823)
* webui: fix --webui-config-file settings not applied on load
* chore: update webui build output | [
{
"path": "tools/server/webui/src/lib/services/parameter-sync.service.ts",
"patch": "@@ -159,6 +159,74 @@ export const SYNCABLE_PARAMETERS: SyncableParameter[] = [\n \t\tserverKey: 'fullHeightCodeBlocks',\n \t\ttype: SyncableParameterType.BOOLEAN,\n \t\tcanSync: true\n+\t},\n+\t{\n+\t\tkey: 'systemMessage',... | 2026-03-23T10:25:35 |
electron/electron | 34129b83a471cbe2aa4de153c1af084a1cd8be83 | b1777c5ad16bd9f270e4558682c99fe7d593bf73 | chore: error catching in node_bindings (#32858) | [
{
"path": "shell/common/node_bindings.cc",
"patch": "@@ -462,8 +462,7 @@ node::Environment* NodeBindings::CreateEnvironment(\n \n args.insert(args.begin() + 1, init_script);\n \n- isolate_data_ =\n- node::CreateIsolateData(context->GetIsolate(), uv_loop_, platform);\n+ isolate_data_ = node::CreateI... | 2022-02-17T05:56:53 |
huggingface/transformers | b9042c4e9a00376412e4b2187161a8af52baa133 | 0e9609c5d4b621dd6ef6b12e9bb73eedaf609e8c | fix: Prevent AutoTokenizer type mismatch from directory name substrin… (#43791)
When saving a tokenizer to a local directory and reloading it, the tokenizer
type could change to an incorrect class (or fall back to TokenizersBackend)
if the directory name contained a model type substring.
We're removing this fallback ... | [
{
"path": "src/transformers/models/auto/configuration_auto.py",
"patch": "@@ -1421,17 +1421,10 @@ def from_pretrained(cls, pretrained_model_name_or_path: str | os.PathLike[str],\n \"`pip install git+https://github.com/huggingface/transformers.git`\"\n )\n retu... | 2026-02-06T15:32:40 |
facebook/react | cd2b79dedd6d81abb0b01d38396afa083feaf9e9 | 6d394e3d26a1e785d0510d732eedcdd2be3e92b5 | Initial (client-only) async actions support (#26621)
Implements initial (client-only) support for async actions behind a
flag. This is an experimental feature and the design isn't completely
finalized but we're getting closer. It will be layered alongside other
features we're working on, so it may not feel complete... | [
{
"path": "packages/react-reconciler/src/ReactFiberAsyncAction.js",
"patch": "@@ -0,0 +1,130 @@\n+/**\n+ * Copyright (c) Meta Platforms, Inc. and affiliates.\n+ *\n+ * This source code is licensed under the MIT license found in the\n+ * LICENSE file in the root directory of this source tree.\n+ *\n+ * @flow... | 2023-04-19T17:33:11 |
ollama/ollama | 9679f40146ed7e148a64abb14c7c619b162c1875 | 3892c3a7032c99db250c3266276c4525d243950a | ml: Allow models to constrain inputs to a single batch
Models may require that a set of inputs all be processed as part
of the same batch. For example, if an image has multiple patches
with fully connected attention between them, we should not split
the batch in the middle of an image.
Fixes #9697 | [
{
"path": "integration/llm_image_test.go",
"patch": "@@ -66,6 +66,35 @@ func TestIntegrationMllama(t *testing.T) {\n \tDoGenerate(ctx, t, client, req, []string{resp}, 240*time.Second, 30*time.Second)\n }\n \n+func TestIntegrationSplitBatch(t *testing.T) {\n+\timage, err := base64.StdEncoding.DecodeString(im... | 2025-03-12T23:56:11 |
ggml-org/llama.cpp | cc18f965b6322513c0ae2ae65420067dd873b360 | 84ffd0c192b120495fd5eb7e922aa7e857304fb2 | fix(openvino): explicit memset in buffer_context allocation (#20857)
* fix(openvino): explicit memset in buffer_context allocation
* minor
---------
Co-authored-by: Dan Hoffman <dhoffman@cyket.net>
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com> | [
{
"path": "ggml/src/ggml-openvino/ggml-openvino.cpp",
"patch": "@@ -97,6 +97,8 @@ struct ggml_backend_openvino_buffer_context {\n ov_buffer = std::make_shared<ov::intel_gpu::ocl::USMTensor>(std::move(usm_tensor));\n } else {\n data = ggml_aligned_malloc(size);\n+ G... | 2026-03-23T06:05:37 |
golang/go | 28f4e14ebe281d8e46cba430bfd123ce21fcd0cc | a2a4f00783701272e4804b7488adef673ef46666 | crypto/internal/fips140deps: fix test for running in FIPS snapshot
In a FIPS snapshot, the import paths have a snapshot version number.
Remove that version in the test before proceeding with the usual checks.
Change-Id: I15c9d11dcac6d33330b334b8e5056c215bffa75c
Reviewed-on: https://go-review.googlesource.com/c/go/+/6... | [
{
"path": "src/crypto/internal/fips140deps/fipsdeps_test.go",
"patch": "@@ -40,17 +40,27 @@ func TestImports(t *testing.T) {\n {{range .XTestImports -}}\n {{$path}} {{.}}\n {{end -}}`, \"crypto/internal/fips140/...\")\n-\tout, err := cmd.CombinedOutput()\n+\tbout, err := cmd.CombinedOutput()\n \tif err != n... | 2024-11-20T14:51:27 |
electron/electron | 600c37160b8811156997d55ab29b2943eb977b7a | 512bb96dcbd085f8f3925350d0e15c2712cc71fa | docs: fix relative link in developer documentation (#32844) | [
{
"path": "docs/development/pull-requests.md",
"patch": "@@ -180,7 +180,7 @@ $ git push origin my-branch\n ### Step 9: Opening the Pull Request\n \n From within GitHub, opening a new pull request will present you with a template\n-that should be filled out. It can be found [here](../../.github/PULL_REQUEST_... | 2022-02-15T21:54:28 |
huggingface/transformers | 0e9609c5d4b621dd6ef6b12e9bb73eedaf609e8c | 5a1016a95e21185542f7c5adebebcbe5419566eb | [`Repo Consistency`] Fix rms norm (#43803)
fix | [
{
"path": "src/transformers/models/vibevoice_acoustic_tokenizer/modeling_vibevoice_acoustic_tokenizer.py",
"patch": "@@ -82,15 +82,15 @@ class VibeVoiceAcousticTokenizerDecoderOutput(ModelOutput):\n \n @use_kernel_forward_from_hub(\"RMSNorm\")\n class VibeVoiceAcousticTokenizerRMSNorm(nn.Module):\n- def ... | 2026-02-06T14:59:18 |
facebook/react | 1f248bdd7199979b050e4040ceecfe72dd977fd1 | b90e8ebaa5b692bcee4198eacb56f9123b62dd10 | Switching checked to null should leave the current value (#26667)
I accidentally made a behavior change in the refactor. It turns out that
when switching off `checked` to an uncontrolled component, we used to
revert to the concept of "initialChecked" which used to be stored on
state.
When there's a diff to this ... | [
{
"path": "packages/react-dom-bindings/src/client/ReactDOMComponent.js",
"patch": "@@ -834,6 +834,7 @@ export function setInitialProperties(\n // listeners still fire for the invalid event.\n listenToNonDelegatedEvent('invalid', domElement);\n \n+ let name = null;\n let type = null;\n... | 2023-04-19T15:46:29 |
ollama/ollama | 65b88c544f08ce3e5b1d193e82b72735095f795c | a422ba39c94adc870da84e5fa442c0bf81c77f27 | fix divide by zero | [
{
"path": "fs/ggml/ggml.go",
"patch": "@@ -579,6 +579,10 @@ func (f GGML) GraphSize(context, batch uint64, kvCacheType string) (kv, partialO\n }\n \n func (llm GGML) VisionGraphSize() (weights, graphSize uint64) {\n+\tif llm.KV().Uint(\"vision.block_count\") == 0 {\n+\t\treturn\n+\t}\n+\n \tfor name, layer ... | 2025-03-13T22:05:42 |
golang/go | b631b8e674195f6cdeca363ccb01279f14b10f13 | a9428eab40f799c96c0de7385545c03d989d491a | go/types, types: better error message position for invalid receiver errors
Errors related to invalid receivers are based on the receiver base type.
Position the error message at the receiver base type, not the receiver
variable.
Add an additional example with an (invalid) generic receiver type.
Also, fix a panic whe... | [
{
"path": "src/cmd/compile/internal/types2/issues_test.go",
"patch": "@@ -852,7 +852,7 @@ import \"C\"\n \n type Layout = C.struct_layout\n \n-func (l /* ERROR \"cannot define new methods on non-local type Layout\" */ *Layout) Binding() {}\n+func (l *Layout /* ERROR \"cannot define new methods on non-local ... | 2024-11-20T21:57:23 |
electron/electron | 4fa3310887490938d35514522eae09157da88663 | ceab6a146a8dc6588c4f34719f2e0e069270eabd | build: add stack_trace.h to main delegate (#32854)
* build: add stack_trace.h to main delegate
* build: trigger circleci
* build: free up a tiny bit more space
* build: disable use_thin_lto
* fixup build/args/all.gn
Co-authored-by: Samuel Attard <sam@electronjs.org>
Co-authored-by: John Kleinschmidt <j... | [
{
"path": ".circleci/build_config.yml",
"patch": "@@ -349,10 +349,10 @@ step-get-more-space-on-mac: &step-get-more-space-on-mac\n command: |\n if [ \"`uname`\" == \"Darwin\" ]; then\n sudo mkdir -p $TMPDIR/del-target\n- if [ \"$TARGET_ARCH\" == \"arm64\" ]; then\n- # Remoun... | 2022-02-11T01:43:14 |
huggingface/transformers | 5a1016a95e21185542f7c5adebebcbe5419566eb | 711f27972f1bdffc3b0f27b21a57416bc43a2a5b | Remove unconditional train_batch_size assignment (#43770)
* Remove unconditional train_batch_size assignment
The train_batch_size should only be saved to TrainerState when
auto_find_batch_size is enabled (which is already handled in the
auto_find_batch_size block at line 2251). The unconditional assignment
caused iss... | [
{
"path": "src/transformers/trainer.py",
"patch": "@@ -2101,7 +2101,10 @@ def train(\n self._load_from_checkpoint(resume_from_checkpoint)\n # In case of repeating the find_executable_batch_size, set `self._train_batch_size` properly\n state = TrainerState.load_from_js... | 2026-02-06T14:46:43 |
ggml-org/llama.cpp | ec2b787ebe975cab1fa592c2505a6d3a9e4ff2e7 | d3ac030a5d1edfbb7f7150126f6b9e9bf7c00c26 | mtmd: Add dynamic high-resolution image preprocessing for InternVL model (#20847)
* added support for internvl's dynamic high-resolution (Qianfan-OCR needed)
* add min/max dynamic patch to gguf meta
* clean up
* simplified handling min/max dynamic patch
* reuse llava_uhd logic for slice images
* provide default v... | [
{
"path": "convert_hf_to_gguf.py",
"patch": "@@ -4273,6 +4273,16 @@ def modify_tensors(self, data_torch: Tensor, name: str, bid: int | None) -> Iter\n \n @ModelBase.register(\"InternVisionModel\")\n class InternVisionModel(MmprojModel):\n+\n+ min_dynamic_tiles: int = 0\n+ max_dynamic_tiles: int = 0\n+... | 2026-03-23T00:06:30 |
facebook/react | a227bcd4f4754b0a3f44f84e1e63a79d0ae130d3 | c6db19f9cdec34bca3625a483a2f85181193b885 | chore[devtools/release-scripts]: update messages / fixed npm view com… (#26660)
Some minor changes, observed while working on 24.7.5 release:
- Updated numeration of text instructions
- `reactjs.org` -> `react.dev`
- Fixed using `npm view` command for node 16+, `publish-release` script
currently fails if used with... | [
{
"path": "packages/react-devtools-extensions/chrome/test.js",
"patch": "@@ -7,7 +7,7 @@ const {resolve} = require('path');\n const {argv} = require('yargs');\n \n const EXTENSION_PATH = resolve('./chrome/build/unpacked');\n-const START_URL = argv.url || 'https://reactjs.org/';\n+const START_URL = argv.url ... | 2023-04-19T09:05:16 |
golang/go | 558f5372fc524e69ade3ab3fe36b1913a4095398 | 60299c251364c6f4a830721e4953230260d118a0 | cmd/compile,testing: implement one-time rampup logic for testing.B.Loop
testing.B.Loop now does its own loop scheduling without interaction with b.N.
b.N will be updated to the actual iterations b.Loop controls when b.Loop returns false.
This CL also added tests for fixed iteration count (benchtime=100x case).
This ... | [
{
"path": "src/cmd/compile/internal/inline/interleaved/interleaved.go",
"patch": "@@ -164,6 +164,20 @@ func fixpoint(fn *ir.Func, match func(ir.Node) bool, edit func(ir.Node) ir.Node)\n \t\t\tif base.Flag.LowerM > 1 {\n \t\t\t\tfmt.Printf(\"%v: skip inlining within testing.B.loop for %v\\n\", ir.Line(n), n)... | 2024-11-13T23:34:51 |
huggingface/transformers | 711f27972f1bdffc3b0f27b21a57416bc43a2a5b | 8c3ac8f1446cb57dd7c6448f5df8478cd9fc22c7 | fix(testing): Fix BLOOM tokenizer, CLAP audio features, and CLVP text tester usage in tests (#43798)
fix(testing): Small test-only fixes for CLAP, BLOOM, and CLVP | [
{
"path": "tests/models/bloom/test_modeling_bloom.py",
"patch": "@@ -544,7 +544,7 @@ def test_batch_generation(self):\n \n input_sentence = [\"I enjoy walking with my cute dog\", \"I enjoy walking with my cute dog\"]\n \n- inputs = tokenizer.batch_encode_plus(input_sentence, return_tensors=\"... | 2026-02-06T13:55:53 |
ggml-org/llama.cpp | d3ac030a5d1edfbb7f7150126f6b9e9bf7c00c26 | 49bfddeca18e62fa3d39114a23e9fcbdf8a22388 | mtmd : fix LightOnOCR image preprocessing (#20877) | [
{
"path": "tools/mtmd/clip.cpp",
"patch": "@@ -1161,7 +1161,6 @@ struct clip_model_loader {\n hparams.set_warmup_n_tokens(16*16);\n } break;\n case PROJECTOR_TYPE_PIXTRAL:\n- case PROJECTOR_TYPE_LIGHTONOCR:\n {\n ... | 2026-03-23T00:04:14 |
facebook/react | a0777212df7009cc26e47ffd46444276d0be923d | fdbfac8f8216839520d093513908918de591468f | Propagate scope declarations to parent scopes
Fix for bug demonstrated in #1506. When we add variables as output of their
defining scope, we need to propagate this information upwards to all parent
scopes which are not current active. | [
{
"path": "compiler/forget/src/ReactiveScopes/PropagateScopeDependencies.ts",
"patch": "@@ -23,6 +23,7 @@ import {\n eachInstructionValueOperand,\n eachPatternOperand,\n } from \"../HIR/visitors\";\n+import { empty, Stack } from \"../Utils/Stack\";\n import { assertExhaustive } from \"../Utils/utils\";\... | 2023-04-19T04:27:11 |
golang/go | b8ba5b440b1f84920e80852ec984520046adaf3a | ae9938fae287606fb08395f4248393ea4acaa071 | cmd/cgo: improve error message for unknown name
Fixes #70472
Change-Id: Ib50d1dd5a648afc199ed62a507d54d0b0ccb1da2
Reviewed-on: https://go-review.googlesource.com/c/go/+/630375
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Auto-Submit: Ian Lance Taylor <iant@golang.org>
Revi... | [
{
"path": "src/cmd/cgo/gcc.go",
"patch": "@@ -490,7 +490,7 @@ func (p *Package) guessKinds(f *File) []*Name {\n \t\t\t\t// Don't report an error, and skip adding n to the needType array.\n \t\t\t\tcontinue\n \t\t\t}\n-\t\t\terror_(f.NamePos[n], \"could not determine kind of name for C.%s\", fixGo(n.Go))\n+\... | 2024-11-20T21:51:53 |
ollama/ollama | 80c7ce381ba5bd435503323007c4049fc1c3faef | ccfd41c4f0f1cbbb57ba19f79a2be36c07825b29 | fix: change default context size for gemma3 (#9744) | [
{
"path": "convert/convert_gemma3.go",
"patch": "@@ -87,7 +87,7 @@ func (p *gemma3Model) KV(t *Tokenizer) ggml.KV {\n \t\tkv[\"gemma3.embedding_length\"] = p.HiddenSize\n \t\tkv[\"gemma3.feed_forward_length\"] = p.IntermediateSize\n \tdefault:\n-\t\tkv[\"gemma3.context_length\"] = cmp.Or(p.MaxPositionEmbedd... | 2025-03-13T20:59:19 |
ggml-org/llama.cpp | bd3f1d9d65113737e795cd6d62eb66f90628663c | 23c9182ce872666aff6bd0fc571c3e3b6ae5fd89 | CUDA: fix BF16 FA compilation (#20865) | [
{
"path": "ggml/src/ggml-cuda/convert.cuh",
"patch": "@@ -42,11 +42,15 @@ template<typename dst_t, typename src_t>\n } else if constexpr(std::is_same_v<src_t, float2> && std::is_same_v<dst_t, half2>) {\n return __float22half2_rn(x);\n } else if constexpr(std::is_same_v<src_t, nv_bfloat162> &... | 2026-03-22T16:53:33 |
electron/electron | 28ada6ea8bd2abeca8812e6d284c1f5f3974286b | e9a43be9beb91a4cb7af04242fe9e27955aa675c | chore: bump chromium to 100.0.4857.0 (main) (#32419)
* chore: bump chromium in DEPS to 99.0.4819.0
* chore: update patches
* chore: bump chromium in DEPS to 99.0.4824.0
* chore: update patches
* chore: bump chromium in DEPS to 99.0.4827.0
* chore: update patches
* 3352511: PiP: Add inkdrop and pointe... | [
{
"path": "BUILD.gn",
"patch": "@@ -472,8 +472,8 @@ source_set(\"electron_lib\") {\n \n if (is_linux) {\n deps += [\n- \"//build/config/linux/gtk:gtkprint\",\n \"//components/crash/content/browser\",\n+ \"//ui/gtk:gtk_config\",\n ]\n }\n \n@@ -551,7 +551,7 @@ source_set(\"electro... | 2022-02-10T02:58:52 |
facebook/react | 96fd2fb726130d2980e6d450f5d9e468f922b8b9 | d8089f2cf27a17872a13805fce415b74fd3b107e | (patch)[DevTools] bug fix: backend injection logic not working for undocked devtools window (#26665)
bugfix for #26492
This bug would cause users unable to use the devtools (component tree
empty).
The else-if logic is broken when user switch to undocked devtools mode
(separate window) because `sender.tab` woul... | [
{
"path": "packages/react-devtools-extensions/src/background.js",
"patch": "@@ -172,6 +172,7 @@ chrome.tabs.onUpdated.addListener((tabId, changeInfo, tab) => {\n \n chrome.runtime.onMessage.addListener((request, sender) => {\n const tab = sender.tab;\n+ // sender.tab.id from content script points to the ... | 2023-04-19T00:39:22 |
huggingface/transformers | 8c3ac8f1446cb57dd7c6448f5df8478cd9fc22c7 | 49dd297930605343d8c498dd7ac026819667fc3c | [`Modular Dependencies`] Fixup qwen rms norms (#43772)
* fix
* type hints | [
{
"path": "src/transformers/models/afmoe/modeling_afmoe.py",
"patch": "@@ -107,15 +107,15 @@ def forward(self, x, position_ids):\n \n @use_kernel_forward_from_hub(\"RMSNorm\")\n class AfmoeRMSNorm(nn.Module):\n- def __init__(self, hidden_size, eps=1e-6):\n+ def __init__(self, hidden_size, eps: float =... | 2026-02-06T12:30:15 |
golang/go | 2e07ff35436bc9adc67885f17cdd9042118d0ffd | 5242bfa26afdd3541b9a70a16431ceb616f5510d | cmd/go: remove base.ShortPathConservative
This CL first removes the base.ShortPathConservative function. It had
two classes of uses. The first was in opening files where the paths end
up in error messages. In all those cases, the non-shortened paths are
used to open the files, and ShortPath is only used for the error
... | [
{
"path": "src/cmd/go/internal/base/path.go",
"patch": "@@ -6,13 +6,12 @@ package base\n \n import (\n \t\"errors\"\n+\t\"io/fs\"\n \t\"os\"\n \t\"path/filepath\"\n \t\"runtime\"\n \t\"strings\"\n \t\"sync\"\n-\n-\t\"cmd/go/internal/str\"\n )\n \n // UncachedCwd returns the current working directory.\n@@ -3... | 2024-11-20T18:39:14 |
ggml-org/llama.cpp | 81bc4d3ddcaa1767b60837d73a0c31babb1190cf | f40a80b4f3cd00c4c405c45b7f316f7e77352323 | server: fix Host header (#20843)
It should include port when it's not default. | [
{
"path": "tools/server/server-models.cpp",
"patch": "@@ -1181,7 +1181,8 @@ server_http_proxy::server_http_proxy(\n continue;\n }\n if (key == \"Host\" || key == \"host\") {\n- req.set_header(key, host);\n+ bool is_default_port = (scheme ... | 2026-03-22T14:29:22 |
Subsets and Splits
Assembly Language GitHub Issues
Retrieves a sample of assembly-language related commits with their details, but doesn't provide meaningful analysis or patterns beyond basic filtering.
Swift Compiler Issues Analysis
Retrieves all training data for the Swift programming language repository, providing basic filtering but offering limited analytical insight beyond identifying relevant code examples.