repo stringclasses 10
values | pr_number int64 7 155k | title stringlengths 4 137 | body stringlengths 0 68.4k | buggy_commit stringlengths 40 40 ⌀ | fix_commit stringlengths 40 40 | buggy_distance int64 1 30 ⌀ | confidence stringclasses 3
values | files listlengths 1 5 |
|---|---|---|---|---|---|---|---|---|
vuejs/vue | 9,621 | [automated] Patreon backers update (Mar 2019 | This is an automated pull request. | 0c4c78bfc026bed79e76d0d0009b6ed15b6bb008 | 19c0f80d2cfaa1327aa019663b2e33dff6be0787 | 7 | medium | [
{
"filename": "BACKERS.md",
"patch": "@@ -319,15 +319,15 @@ Funds donated via Patreon go directly to support Evan You's full-time work on Vu\n <h2 align=\"center\">Generous Backers via Patreon ($50+)</h2>\n \n <!--50 start-->\n-- Jamie K\n - Wasim Khamlichi\n - errorrik\n - Alex Balashov\n - Konstantin Levi... |
facebook/react | 33,660 | [Fiber] Highlight a Component with Deeply Equal Props in the Performance Track | Stacked on #33658 and #33659.
If we detect that a component is receiving only deeply equal objects, then we highlight it as potentially problematic and worth looking into.
<img width="1055" alt="Screenshot 2025-06-27 at 4 15 28 PM" src="https://github.com/user-attachments/assets/e96c6a05-7fff-4fd7-b59a-36ed79f8e6... | 508f7aa78ff53d058ee1151505efd5c4a4aefa01 | 0b78161d7d76b7fb9786f25dd222010b9e417191 | 3 | medium | [
{
"filename": "packages/react-reconciler/src/ReactFiberCommitWork.js",
"patch": "@@ -143,6 +143,8 @@ import {\n logComponentUnmount,\n logComponentReappeared,\n logComponentDisappeared,\n+ pushDeepEquality,\n+ popDeepEquality,\n } from './ReactFiberPerformanceTrack';\n import {ConcurrentMode, NoMode... |
electron/electron | 49,577 | build(deps): bump lodash from 4.17.21 to 4.17.23 in the npm_and_yarn group across 1 directory | Bumps the npm_and_yarn group with 1 update in the / directory: [lodash](https://github.com/lodash/lodash).
Updates `lodash` from 4.17.21 to 4.17.23
<details>
<summary>Commits</summary>
<ul>
<li><a href="https://github.com/lodash/lodash/commit/dec55b7a3b382da075e2eac90089b4cd00a26cbb"><code>dec55b7</code></a> Bump main... | 3d76d3a469d6fb22f0b6e137e210c81ae9e739fa | c3428fa4133f66d0da336849a71667fb05fe64b8 | 18 | medium | [
{
"filename": "yarn.lock",
"patch": "@@ -6875,16 +6875,16 @@ __metadata:\n linkType: hard\n \n \"glob@npm:^7.1.3, glob@npm:^7.1.6\":\n- version: 7.2.0\n- resolution: \"glob@npm:7.2.0\"\n+ version: 7.2.3\n+ resolution: \"glob@npm:7.2.3\"\n dependencies:\n fs.realpath: \"npm:^1.0.0\"\n infligh... |
ollama/ollama | 10,018 | ml: Remove Output from Context interface | Model implementations should use Input for all of their tensors supplied to the model. This includes tensors that relate to the outputs, which is confusing since there is also an Output funciton.
Since Output is only used internally in GGML and not used by any model implementations, we can remove it from the interfa... | 74bd09652d69c77a4bed34b3afda74c87295115b | 01aa7887221e7bd286ebcb14a088c94ba1c22a99 | 12 | medium | [
{
"filename": "kvcache/causal_test.go",
"patch": "@@ -362,7 +362,6 @@ func (c *testContext) FromIntSlice(s []int32, shape ...int) (ml.Tensor, error) {\n }\n \n func (c *testContext) Input() ml.Context { return c }\n-func (c *testContext) Output() ml.Context { return c }\n func (c *testContext) Layer(in... |
vercel/next.js | 88,985 | [Reapply] Add `useEffectEvent` to disallowed React APIs in Server Components | <!-- Thanks for opening a PR! Your contribution is much appreciated.
To make sure your PR is handled as smoothly as possible we request that you follow the checklist sections below.
Choose the right checklist for the change(s) that you're making:
## For Contributors
### Improving Documentation
- Run `pnpm pr... | null | 7eb2ea491ea89ea753eb2e636edf87bad67574f2 | null | low | [
{
"filename": "crates/next-custom-transforms/src/transforms/react_server_components.rs",
"patch": "@@ -633,6 +633,7 @@ impl ReactServerComponentValidator {\n \"PureComponent\",\n \"useDeferredValue\",\n \"useEffect\",\n+ ... |
facebook/react | 33,499 | Disable ScrollTimeline in Safari | Stacked on #33501.
This disables the use of ScrollTimeline when detected in Safari in the recommended SwipeRecognizer approach. I'm instead using a polyfill using touch events on iOS.
Safari seems set to [release ScrollTimeline soon](https://webkit.org/blog/16993/news-from-wwdc25-web-technology-coming-this-fall-i... | fc41c24aa6e674319aed5bd0b25ca6fb92c268df | dcf83f7c2db2de64ec370df73294cedb658bd896 | 5 | medium | [
{
"filename": "fixtures/view-transition/src/components/SwipeRecognizer.js",
"patch": "@@ -6,6 +6,14 @@ import React, {\n } from 'react';\n \n import ScrollTimelinePolyfill from 'animation-timelines/scroll-timeline';\n+import TouchPanTimeline from 'animation-timelines/touch-pan-timeline';\n+\n+const ua = typ... |
ollama/ollama | 9,909 | Adds molbal/orca-cli to the Community Integrations section in the readme | dd66712e3159161c1de9c39a12fb83edf8813d39 | e5d84fb90b21d71f8eb816656ca0b34191425216 | 1 | high | [
{
"filename": "README.md",
"patch": "@@ -435,6 +435,7 @@ See the [API documentation](./docs/api.md) for all endpoints.\n - [aichat](https://github.com/sigoden/aichat) All-in-one LLM CLI tool featuring Shell Assistant, Chat-REPL, RAG, AI tools & agents, with access to OpenAI, Claude, Gemini, Ollama, Groq, an... | |
facebook/react | 33,662 | [Flight] Use a heuristic to extract a useful description of I/O from the Promise value | It's useful to be able to distinguish between different invocations of common helper libraries (like fetch) without having to click through each one.
This adds a heuristic to extract a useful description of I/O from the Promise value. We try to find things like getUser(id) -> User where User.id is the id or fetch(ur... | e104795f635e2bb423fe9264b1cfdfff15f551e5 | 94fce500bcd314ea19c5c89c6385be6e62397c09 | 2 | medium | [
{
"filename": "packages/react-client/src/ReactFlightPerformanceTrack.js",
"patch": "@@ -300,6 +300,125 @@ function getIOColor(\n }\n }\n \n+function getIODescription(value: any): string {\n+ if (!__DEV__) {\n+ return '';\n+ }\n+ try {\n+ switch (typeof value) {\n+ case 'object':\n+ //... |
nodejs/node | 61,407 | src: use C++ nullptr in webstorage | Using the C-style constant `0` to represent a null pointer is less type safe than the C++-style constant `nullptr`, so we should always be using `nullptr` instead of `0`.
<!--
Before submitting a pull request, please read:
- the CONTRIBUTING guide at https://github.com/nodejs/node/blob/HEAD/CONTRIBUTING.md
- th... | null | 58f5da22237e4e1378b1d45f9a17c935be659763 | null | low | [
{
"filename": "src/node_webstorage.cc",
"patch": "@@ -170,14 +170,17 @@ Maybe<void> Storage::Open() {\n \n int r = sqlite3_open(location_.c_str(), &db);\n CHECK_ERROR_OR_THROW(env(), r, SQLITE_OK, Nothing<void>());\n- r = sqlite3_exec(db, init_sql_v0.data(), 0, 0, nullptr);\n+ r = sqlite3_exec(db, ini... |
huggingface/transformers | 43,432 | Make `pr-repo-consistency-bot.yml` run complete list of fixes | # What does this PR do?
Happy for everyone if we can have such complete fixes from the bot instead of just `check_copies`. | d6edf57ed28e3dee0194fe5bd185e57be0525ee6 | a1306676e920c58ec0818059d59a71aed57f9f60 | 30 | medium | [
{
"filename": ".github/workflows/pr-repo-consistency-bot.yml",
"patch": "@@ -148,12 +148,39 @@ jobs:\n - name: Run checks with trusted script\n id: run_checks\n run: |\n- # Copy trusted script to PR directory\n+ # Security: Copy trusted scripts from main branch and ru... |
ggml-org/llama.cpp | 20,586 | CUDA: limit number of FA stream-k CUDA blocks | On master the CUDA mma FA kernel can launch superfluous CUDA blocks that do not do any useful work but cause overhead. This can happen when running small models on GPUs with many streaming multiprocessors at low batch sizes. This PR fixes this by limiting the number of CUDA blocks to the number that can do useful work.... | null | ae40cd27c85aa30b9cd56033da1d6a954290f7ea | null | low | [
{
"filename": "ggml/src/ggml-cuda/fattn-common.cuh",
"patch": "@@ -892,7 +892,7 @@ void launch_fattn(\n const int ntiles_x = ((Q->ne[1] + ncols1 - 1) / ncols1);\n const int gqa_ratio = Q->ne[2] / K->ne[2];\n const int ntiles_z_gqa = ((gqa_ratio + ncols2 - 1) / ncols2);\n- const int nti... |
facebook/react | 33,659 | [Fiber] Switch back to using performance.measure for trigger logs | Stacked on #33658.
Unfortunately `console.timeStamp` has the same bug that `performance.measure` used to have where equal start/end times stack in call order instead of reverse call-order. We rely on that in general so we should really switch back all.
But there is one case in particular where we always add the s... | e104795f635e2bb423fe9264b1cfdfff15f551e5 | 508f7aa78ff53d058ee1151505efd5c4a4aefa01 | 1 | medium | [
{
"filename": "packages/react-reconciler/src/ReactFiberPerformanceTrack.js",
"patch": "@@ -101,28 +101,22 @@ function logComponentTrigger(\n trigger: string,\n ) {\n if (supportsUserTiming) {\n+ reusableComponentOptions.start = startTime;\n+ reusableComponentOptions.end = endTime;\n+ reusableCo... |
ollama/ollama | 9,879 | Add Ollamb to the list of Ollama clients | This PR adds OLLAMB to the list of Ollama clients in the README.
OLLAMB is a simple yet rich-in-features, cross-platform client built with Flutter and designed for Ollama. It offers an easy installation process and a smooth user experience.
<img width="1552" alt="OLLAMB" src="https://github.com/user-attachments/a... | 131f0355a59f4840b057fb8f3c2e59e456f91041 | dd66712e3159161c1de9c39a12fb83edf8813d39 | 5 | medium | [
{
"filename": "README.md",
"patch": "@@ -394,6 +394,7 @@ See the [API documentation](./docs/api.md) for all endpoints.\n - [Reins](https://github.com/ibrahimcetin/reins) (Easily tweak parameters, customize system prompts per chat, and enhance your AI experiments with reasoning model support.)\n - [Ellama](h... |
electron/electron | 49,629 | build(deps): bump github/codeql-action from 4.32.0 to 4.32.1 | Bumps [github/codeql-action](https://github.com/github/codeql-action) from 4.32.0 to 4.32.1.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a href="https://github.com/github/codeql-action/releases">github/codeql-action's releases</a>.</em></p>
<blockquote>
<h2>v4.32.1</h2>
<ul>
<li>A warning is now sho... | 441729c3a0b6f546382f52554e8891dbb4476a96 | bdf2b674622fac64112f4ea7299120ab5a332dd0 | 21 | medium | [
{
"filename": ".github/workflows/scorecards.yml",
"patch": "@@ -51,6 +51,6 @@ jobs:\n \n # Upload the results to GitHub's code scanning dashboard.\n - name: \"Upload to code-scanning\"\n- uses: github/codeql-action/upload-sarif@b20883b0cd1f46c72ae0ba6d1090936928f9fa30 # v3.29.5\n+ ... |
ggml-org/llama.cpp | 20,595 | ggml: avoid creating CUDA context during device init | *Make sure to read the [contributing guidelines](https://github.com/ggml-org/llama.cpp/blob/master/CONTRIBUTING.md) before submitting a PR*
ggml_cuda_init() calls cudaSetDevice() on every GPU just to query free VRAM for logging. This triggers the creation of a CUDA primary context (120-550 MB depending on GPU), whic... | 8b7d340b6ffbb1bc47093f5b4b840bc2ff996a75 | ceef6b5233c3b31f454632c48fb42af16944bc5b | 2 | high | [
{
"filename": "ggml/src/ggml-cuda/ggml-cuda.cu",
"patch": "@@ -254,11 +254,6 @@ static ggml_cuda_device_info ggml_cuda_init() {\n info.devices[id].supports_cooperative_launch = false;\n #endif // !(GGML_USE_MUSA)\n \n- // cudaMemGetInfo returns info for the current device\n- size_t fre... |
ollama/ollama | 9,987 | Improve memory estimates for sliding window attention | Recent optimizations for sliding window attention significantly reduced memory usage for models that use it. However, our memory estimates didn't reflect that, causing us to spill over to system memory unnecessarily in cases where GPU memory is constrained. | f4f0992b6ea5d651eff609461c24ece936bd5708 | f66216e3990b73869341c58ac9561b26c468c558 | 1 | medium | [
{
"filename": "fs/ggml/ggml.go",
"patch": "@@ -413,7 +413,7 @@ func Decode(rs io.ReadSeeker, maxArraySize int) (*GGML, int64, error) {\n \t}, offset, nil\n }\n \n-func (f GGML) GraphSize(context, batch uint64, kvCacheType string) (kv, partialOffload, fullOffload uint64) {\n+func (f GGML) GraphSize(context, ... |
vuejs/vue | 9,647 | fix(scheduler): getNow detection can randomly fail (fix #9632) | The previous detection code compared time stamps based on Date.now()
which are not monotonic, so the check could fail due to clock skew or
adjustments.
This fix changes the check to compare against performance.now() if it is
supported, because it is monotonic (strictly increasing).
<!--
Please make sure to re... | null | da77d6a98bdccd8a2c8bfdfe6b9cb46efcb1193c | null | low | [
{
"filename": "src/core/observer/scheduler.js",
"patch": "@@ -47,10 +47,15 @@ let getNow: () => number = Date.now\n // timestamp can either be hi-res (relative to page load) or low-res\n // (relative to UNIX epoch), so in order to compare time we have to use the\n // same timestamp type when saving the flus... |
facebook/react | 33,658 | [Fiber] Show Diff Render Props in Performance Track in DEV | <img width="634" alt="Screenshot 2025-06-27 at 1 13 20 PM" src="https://github.com/user-attachments/assets/dc8c488b-4a23-453f-918f-36b245364934" />
We have to be careful with performance in DEV. It can slow down DX since these are ran whether you're currently running a performance trace or not. It can also show up a... | e9cab42ece435ac3478ec85847e352177e596ae0 | e104795f635e2bb423fe9264b1cfdfff15f551e5 | 10 | medium | [
{
"filename": "packages/react-client/src/ReactFlightPerformanceTrack.js",
"patch": "@@ -103,10 +103,10 @@ export function logComponentRender(\n if (__DEV__ && debugTask) {\n const properties: Array<[string, string]> = [];\n if (componentInfo.key != null) {\n- addValueToProperties('key... |
facebook/react | 33,576 | Clear width/height from Keyframes to Optimize View Transitions | View Transitions has this annoying quirk where it adds `width` and `height` to keyframes automatically when generating keyframes even when it's not needed. This causes them to deopt from running on the compositor thread in both Chrome and Safari. @bramus has a [good article on it](https://www.bram.us/2025/02/07/view-tr... | fc41c24aa6e674319aed5bd0b25ca6fb92c268df | c0d151ce7ed9ebdcbcb7ea69c31d68b03b852d37 | 1 | medium | [
{
"filename": "packages/react-dom-bindings/src/client/ReactFiberConfigDOM.js",
"patch": "@@ -2133,6 +2133,84 @@ export function startViewTransition(\n });\n // $FlowFixMe[prop-missing]\n ownerDocument.__reactViewTransition = transition;\n+\n+ const readyCallback = () => {\n+ const docume... |
vercel/next.js | 88,961 | feat(next-codemod): add agents-md command for AI coding agents | ## Summary
Adds a new `agents-md` subcommand to `@next/codemod` that generates a Next.js documentation index for AI coding agents (Claude, Cursor, GitHub Copilot, etc.).
## Usage
```bash
# Interactive mode - prompts for version and target file
npx @next/codemod agents-md
# Non-interactive mode
npx @next/codemod age... | null | 1e859c06aafad02037e6240855c61c5968c2d7ba | null | low | [
{
"filename": "packages/next-codemod/.gitignore",
"patch": "@@ -2,4 +2,5 @@\n *.js\n *.js.map\n !transforms/__tests__/**/*.js\n-!transforms/__testfixtures__/**/*.js\n\\ No newline at end of file\n+!transforms/__testfixtures__/**/*.js\n+!lib/__tests__/**/*.js\n\\ No newline at end of file",
"additions": ... |
vercel/next.js | 88,889 | Turbopack: change invalidator and immutable to data category | ### What?
save a few bits | 9e97759ab8fc876dca1fbfc3e9746d34d113b0d8 | cedfb53ae241af4540c8597a6c7ffdeb7e9fdf80 | 1 | medium | [
{
"filename": "turbopack/crates/turbo-tasks-backend/src/backend/mod.rs",
"patch": "@@ -2091,7 +2091,7 @@ impl<B: BackingStorage> TurboTasksBackendInner<B> {\n if let Some(dependencies) = task_dependencies_for_immutable\n && dependencies\n .iter()\n- .all(|&... |
facebook/react | 32,722 | Add compareDocumentPosition to fragment instances | This adds `compareDocumentPosition(otherNode)` to fragment instances.
The semantics implemented are meant to match typical element positioning, with some fragment specifics. See the unit tests for all expectations.
- An element preceding a fragment is `Node.DOCUMENT_POSITION_PRECEDING`
- An element after a fragm... | 54a50729cc47a884c2110d7c59dd5f850748e142 | e5a8de81e57181692d33ce916dfd6aa23638ec92 | 3 | medium | [
{
"filename": "packages/react-dom-bindings/src/client/ReactFiberConfigDOM.js",
"patch": "@@ -37,6 +37,11 @@ import {runWithFiberInDEV} from 'react-reconciler/src/ReactCurrentFiber';\n import hasOwnProperty from 'shared/hasOwnProperty';\n import {checkAttributeStringCoercion} from 'shared/CheckStringCoercion... |
nodejs/node | 60,595 | events: remove redundant todo | As [said by the person who put the todo in the first place](https://github.com/nodejs/node/pull/55312#discussion_r1798472632), it has been redundant for a while now
We have `kWeakHandler: this` now
PR removes it to prevent confusion | null | 5f7c3a6d1c31b7a098cf15bdaed3defb07bc0f10 | null | low | [
{
"filename": "lib/internal/event_target.js",
"patch": "@@ -639,8 +639,6 @@ class EventTarget {\n if (signal.aborted) {\n return;\n }\n- // TODO(benjamingr) make this weak somehow? ideally the signal would\n- // not prevent the event target from GC.\n signal.a... |
huggingface/transformers | 43,418 | Fix in-place modification of inputs_embeds in Kosmos-2.5 forward | # What does this PR do?
Fixes a PyTorch autograd error when fine-tuning `Kosmos2_5ForConditionalGeneration` with PEFT (LoRA).
The model forward pass performed an in-place assignment on `inputs_embeds`, which can be a leaf tensor with `requires_grad=True` during training. This caused the following runtime error:
... | null | 95be7ef5ed51cd067f1eb3991ae7e186a95d2100 | null | low | [
{
"filename": "src/transformers/models/kosmos2_5/modeling_kosmos2_5.py",
"patch": "@@ -1091,8 +1091,9 @@ def forward(\n \n # Ignore copy\n if image_embeds is not None:\n+ inputs_embeds = inputs_embeds.clone()\n inputs_embeds[image_embeds_position_mask == 1] = image_emb... |
vercel/next.js | 88,997 | Improve agents-md prompt to force doc retrieval | ## Summary
Updates the instruction in `agents-md` [CLAUDE.md](http://CLAUDE.md) output to force LLMs to actually read the docs instead of relying on stale pre-training knowledge.
**Before:**
```
IMPORTANT: Prefer retrieval-led reasoning over pre-training-led reasoning for any Next.js tasks.
```
**After:**
```
STOP... | 1e859c06aafad02037e6240855c61c5968c2d7ba | 39ce0124e4af8c32c5bf8dae05c801fac17a2954 | 4 | medium | [
{
"filename": "packages/next-codemod/lib/agents-md.ts",
"patch": "@@ -10,7 +10,7 @@ import fs from 'fs'\n import path from 'path'\n import os from 'os'\n \n-export interface NextjsVersionResult {\n+interface NextjsVersionResult {\n version: string | null\n error?: string\n }\n@@ -68,13 +68,13 @@ functio... |
ggml-org/llama.cpp | 20,551 | vulkan: use graphics queue on AMD | I'm not sure why, but the graphics queue is slightly faster in tg on AMD than the compute queue, and this also fixes the partial offload issue I fixed in #19976, so the second queue no longer has to be enabled by default. I got the idea from @zedbytes reporting that tg goes up when running with `RADV_DEBUG=nocompute`.
... | null | 1a3d8edbbaba7f6e36096982c7c8a7ce11f4a7e8 | null | low | [
{
"filename": "ggml/src/ggml-vulkan/ggml-vulkan.cpp",
"patch": "@@ -4981,8 +4981,10 @@ static vk_device ggml_vk_get_device(size_t idx) {\n std::vector<vk::QueueFamilyProperties> queue_family_props = device->physical_device.getQueueFamilyProperties();\n \n // Try to find a non-graphics comput... |
vuejs/vue | 9,653 | Fix #9641: v-bind object should be overridable by single bindings | <!--
Please make sure to read the Pull Request Guidelines:
https://github.com/vuejs/vue/blob/dev/.github/CONTRIBUTING.md#pull-request-guidelines
-->
<!-- PULL REQUEST TEMPLATE -->
<!-- (Update "[ ]" to "[x]" to check a box) -->
**What kind of change does this PR introduce?** (check at least one)
- [x] Bugf... | null | 0b57380f10986c6b07e3c240acc06bfd2eddfd1b | null | low | [
{
"filename": "src/core/instance/render-helpers/bind-object-props.js",
"patch": "@@ -7,7 +7,8 @@ import {\n isObject,\n toObject,\n isReservedAttribute,\n- camelize\n+ camelize,\n+ hyphenate\n } from 'core/util/index'\n \n /**\n@@ -45,12 +46,13 @@ export function bindObjectProps (\n : d... |
ollama/ollama | 9,852 | Added useful args to linux command | Added --follow and --pager-end to monitor the output as well as jump to the end of the output. | null | 5e0b904e887fc648fb8a3a55283f8f33063a78eb | null | low | [
{
"filename": "docs/troubleshooting.md",
"patch": "@@ -9,7 +9,7 @@ cat ~/.ollama/logs/server.log\n On **Linux** systems with systemd, the logs can be found with this command:\n \n ```shell\n-journalctl -u ollama --no-pager\n+journalctl -u ollama --no-pager --follow --pager-end \n ```\n \n When you run Ollam... |
electron/electron | 49,615 | fix: handle out of order recording errors in skia graphite | Backport of #49608
See that PR for details.
Notes: Fixed rendering glitches on macOS. | null | faa40332ad9628adc51624dc7789b55d9619dccc | null | low | [
{
"filename": "patches/chromium/.patches",
"patch": "@@ -147,3 +147,4 @@ viz_fix_visual_artifacts_due_to_resizing_root_render_pass_with_dcomp.patch\n viz_do_not_overallocate_surface_on_initial_render.patch\n viz_create_isbufferqueuesupportedandenabled.patch\n viz_fix_visual_artifacts_while_resizing_window_w... |
facebook/react | 33,675 | [Flight] Don't assume _debugStack and _owner is defined for prod elements | We generally treat these types of fields as optional on ReactDebugInfo and should on ReactElement too.
That way we can consume prod payloads from third parties. | 3cfcdfb30720a5b8de0e981c8fdabec1abb61588 | 7216c0f002222cdee3075410f7432d64724640cc | 5 | medium | [
{
"filename": "packages/react-client/src/ReactFlightClient.js",
"patch": "@@ -928,8 +928,8 @@ function createElement(\n type: mixed,\n key: mixed,\n props: mixed,\n- owner: null | ReactComponentInfo, // DEV-only\n- stack: null | ReactStackTrace, // DEV-only\n+ owner: ?ReactComponentInfo, // DEV-onl... |
huggingface/transformers | 43,207 | Fix Mamba2ForCausalLM weight tying | ## What does this PR do?
Fixes #43206
Adds the `_tied_weights_keys` mapping to `Mamba2ForCausalLM` to enable proper weight tying when `tie_word_embeddings=True`.
## The Bug
When `tie_word_embeddings=True`, the embedding weights should be shared with the `lm_head`. However, `Mamba2ForCausalLM` had:
```python
_tied_w... | eff263cd595bb47013af292ba619447cc842f4ab | 10e97cd508218546ef681a2c9b4c519ac0d927c3 | 12 | medium | [
{
"filename": "src/transformers/models/mamba2/modeling_mamba2.py",
"patch": "@@ -930,7 +930,7 @@ def forward(\n \"\"\"\n )\n class Mamba2ForCausalLM(Mamba2PreTrainedModel, GenerationMixin):\n- _tied_weights_keys = {}\n+ _tied_weights_keys = {\"lm_head.weight\": \"backbone.embeddings.weight\"}\n \n... |
facebook/react | 33,673 | Back out "Remove Dead Code in WWW JS" |
Original commit changeset: 65c4decb565b
This was removed by dead code removal. Adding back the TODO with commented out code.
| 65c4decb565b4eb1423518e76dbda7bc40a01c04 | 6a3d16ca74acd28632dabccc142d42ce411a811b | 1 | high | [
{
"filename": "packages/react-server/src/ReactFizzServer.js",
"patch": "@@ -4661,6 +4661,12 @@ function abortTask(task: Task, request: Request, error: mixed): void {\n if (node !== null && typeof node === 'object') {\n // Push a fake component stack frame that represents the await.\n pushHal... |
ggml-org/llama.cpp | 20,578 | vendor : update cpp-httplib to 0.38.0 | 77e20cc1076f351f6ba37b545a20b21c6207cbe5 | 07c6a59b4fc9f6163aa3a4cdc6c4b83646478337 | 25 | medium | [
{
"filename": "scripts/sync_vendor.py",
"patch": "@@ -5,7 +5,7 @@\n import sys\n import subprocess\n \n-HTTPLIB_VERSION = \"refs/tags/v0.37.2\"\n+HTTPLIB_VERSION = \"refs/tags/v0.38.0\"\n \n vendor = {\n \"https://github.com/nlohmann/json/releases/latest/download/json.hpp\": \"vendor/nlohmann/json.h... | |
vuejs/vue | 8,845 | fix #8842: Override v-bind object with kebab-cased props | <!--
Please make sure to read the Pull Request Guidelines:
https://github.com/vuejs/vue/blob/dev/.github/CONTRIBUTING.md#pull-request-guidelines
-->
<!-- PULL REQUEST TEMPLATE -->
<!-- (Update "[ ]" to "[x]" to check a box) -->
**What kind of change does this PR introduce?** (check at least one)
- [x] Bugf... | null | 758524134e71ae025238e16a4c1f2b30a1310fe8 | null | low | [
{
"filename": "src/core/instance/render-helpers/bind-object-props.js",
"patch": "@@ -6,7 +6,8 @@ import {\n warn,\n isObject,\n toObject,\n- isReservedAttribute\n+ isReservedAttribute,\n+ camelize\n } from 'core/util/index'\n \n /**\n@@ -43,12 +44,13 @@ export function bindObjectProps (\n ... |
electron/electron | 49,608 | fix: handle out of order recording errors in skia graphite | #### Description of Change
Refs https://issues.chromium.org/issues/458722690
[Skia graphite](https://blog.chromium.org/2025/07/introducing-skia-graphite-chromes.html) is the new rasterization backend currently enabled on macOS since M135.
There are scenarios where the recorder will go into unrecoverable mode w... | null | db5606b4b357aa6489cd9d1239b12fdf60888afc | null | low | [
{
"filename": "patches/chromium/.patches",
"patch": "@@ -144,3 +144,4 @@ fix_linux_tray_id.patch\n expose_gtk_ui_platform_field.patch\n fix_os_crypt_async_cookie_encryption.patch\n patch_osr_control_screen_info.patch\n+graphite_handle_out_of_order_recording_errors.patch",
"additions": 1,
"deletions"... |
ollama/ollama | 9,907 | [readme]: Add ollama-d library | doc: https://ollama-d.dpldocs.dlang.org/ollama.html
pkg-info: https://code.dlang.org/packages/ollama-d
cc: @BruceMacD | e27e4a3c1b9cb3a34178d5476debf805910e990a | 131f0355a59f4840b057fb8f3c2e59e456f91041 | 30 | medium | [
{
"filename": "README.md",
"patch": "@@ -512,6 +512,7 @@ See the [API documentation](./docs/api.md) for all endpoints.\n - [Ollama for Zig](https://github.com/dravenk/ollama-zig)\n - [Abso](https://github.com/lunary-ai/abso) (OpenAI-compatible TypeScript SDK for any LLM provider)\n - [Nichey](https://github... |
nodejs/node | 57,810 | zlib: reduce code duplication | The offset in the allocated memory was calculated in alloc and free, this makes it a single constant so it only needs to be defined once.
As requested by @Flarna in https://github.com/nodejs/node/pull/57727 | null | 54c456702c7fe6b9b0ef026f864dc1bc9590c88e | null | low | [
{
"filename": "src/node_zlib.cc",
"patch": "@@ -612,9 +612,11 @@ class CompressionStream : public AsyncWrap, public ThreadPoolWork {\n return AllocForBrotli(data, real_size);\n }\n \n+ static constexpr size_t reserveSizeAndAlign =\n+ std::max(sizeof(size_t), alignof(max_align_t));\n+\n static ... |
vercel/next.js | 88,975 | Update font data | This auto-generated PR updates font data with latest available | null | 89cddc26da7627d668a6469c1a36d87b3bc3b3b2 | null | low | [
{
"filename": "packages/font/src/google/font-data.json",
"patch": "@@ -7762,6 +7762,11 @@\n \"styles\": [\"normal\"],\n \"subsets\": [\"cyrillic\", \"cyrillic-ext\", \"devanagari\", \"latin\", \"latin-ext\"]\n },\n+ \"LINE Seed JP\": {\n+ \"weights\": [\"100\", \"400\", \"700\", \"800\"],\n+ ... |
facebook/react | 33,672 | Remove Dead Code in WWW JS | Reviewed By: kassens
Differential Revision: D77531947
| 31d91651e042e4939021f21a3d8799c13684a84b | 65c4decb565b4eb1423518e76dbda7bc40a01c04 | 11 | medium | [
{
"filename": "packages/react-server/src/ReactFizzServer.js",
"patch": "@@ -4661,10 +4661,6 @@ function abortTask(task: Task, request: Request, error: mixed): void {\n if (node !== null && typeof node === 'object') {\n // Push a fake component stack frame that represents the await.\n pushHal... |
ollama/ollama | 9,931 | ollama: fix file descriptor management in Pull | 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 to close each chunked writer as soon as its downloads finish.
Also rename '... | c794fef2f27d141393064665d2774b341d091393 | ce929984a33230269905e0e3cfa335cb8d6ba781 | 4 | medium | [
{
"filename": "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 \t... |
ggml-org/llama.cpp | 20,536 | ggml/hip: fix APU compatibility - soft error handling for hipMemAdviseSetCoarseGrain | Description:
## Problem
On AMD APU/iGPU devices (unified memory architecture, e.g. AMD Strix Halo gfx1151), `hipMemAdviseSetCoarseGrain` returns
`hipErrorInvalidValue` because this hint is not applicable to UMA systems. The current code wraps this call in `CUDA_CHECK()`, which treats
it as a fatal error ... | f2c0dfb7394b3abb5a5afd1c2a94f621bb64236f | 8b7d340b6ffbb1bc47093f5b4b840bc2ff996a75 | 22 | medium | [
{
"filename": "ggml/src/ggml-cuda/ggml-cuda.cu",
"patch": "@@ -124,7 +124,10 @@ static cudaError_t ggml_cuda_device_malloc(void ** ptr, size_t size, int device)\n err = cudaMallocManaged(ptr, size);\n #if defined(GGML_USE_HIP)\n if (err == hipSuccess) {\n- CUDA_CHECK(cudaMemAdvise... |
vuejs/vue | 9,624 | Fixed `RenderContext.scopedSlots` types bug | <!--
Please make sure to read the Pull Request Guidelines:
https://github.com/vuejs/vue/blob/dev/.github/CONTRIBUTING.md#pull-request-guidelines
-->
<!-- PULL REQUEST TEMPLATE -->
<!-- (Update "[ ]" to "[x]" to check a box) -->
**What kind of change does this PR introduce?** (check at least one)
- [x] Bugf... | 241eea19a64550bfdb3f9d7e4197127997572842 | 731e4d0ebc997a6c2b1fc500b08924ff7ba091b6 | 12 | medium | [
{
"filename": "types/options.d.ts",
"patch": "@@ -1,5 +1,5 @@\n import { Vue, CreateElement, CombinedVueInstance } from \"./vue\";\n-import { VNode, VNodeData, VNodeDirective, ScopedSlot } from \"./vnode\";\n+import { VNode, VNodeData, VNodeDirective, NormalizedScopedSlot } from \"./vnode\";\n \n type Const... |
ollama/ollama | 9,897 | ml/backend/ggml: load tensors in 128KiB chunks | also set a limit for number of concurrent goroutines for loading tensors. this change also pipes the progress back to the parent process | 3ed7ad3ab32b458aa2fdb8d0144c546efdb26a72 | 4b34930a31ceb9cc10d95b8bcd60c319f47d8043 | 6 | medium | [
{
"filename": "ml/backend.go",
"patch": "@@ -2,6 +2,7 @@ package ml\n \n import (\n \t\"bytes\"\n+\t\"context\"\n \t\"encoding/binary\"\n \t\"fmt\"\n \t\"os\"\n@@ -80,19 +81,19 @@ type BackendParams struct {\n \tFlashAttention bool\n }\n \n-var backends = make(map[string]func(*os.File, BackendParams) (Backe... |
huggingface/transformers | 43,017 | 🚨Fix default interpolation for PVT to match original implementation (Bicubic) | # What does this PR do?
This PR updates the default interpolation method for `PvtImageProcessor` (both standard and fast versions) from `BILINEAR` to `BICUBIC`.
# Motivation
The original PVT implementation follows the DeiT training recipe, which relies on Bicubic interpolation.
Additionally, the `timm` librar... | null | ef2de73bf21fcf3120ecdf45973aef198365c79a | null | low | [
{
"filename": "src/transformers/models/pvt/image_processing_pvt.py",
"patch": "@@ -47,7 +47,7 @@ class PvtImageProcessor(BaseImageProcessor):\n size (`dict`, *optional*, defaults to `{\"height\": 224, \"width\": 224}`):\n Size of the output image after resizing. Can be overridden by the ... |
electron/electron | 49,594 | fix: issues with xdg activation on Linux | Backport of #49499
See that PR for details.
Notes: Fixed an issue where some `shell` functionality may not work on Linux if the target window is already open.
| 86196dc5886b4023141b2e12d408dc1367ef5fb9 | 58f4af46360584f99565c7c662b8a243f42ad1a1 | 17 | medium | [
{
"filename": "shell/common/platform_util_linux.cc",
"patch": "@@ -7,7 +7,9 @@\n #include <fcntl.h>\n \n #include <stdio.h>\n+#include <memory>\n #include <optional>\n+#include <queue>\n #include <string>\n #include <vector>\n \n@@ -27,8 +29,13 @@\n #include \"base/run_loop.h\"\n #include \"base/strings/esc... |
facebook/react | 33,671 | Align `AttributeConfiguration` type in `ReactNativeTypes` | <!--
Thanks for submitting a pull request!
We appreciate you spending the time to work on these changes. Please provide enough information so that others can review your pull request. The three fields below are mandatory.
Before submitting a pull request, please make sure the following is done:
1. Fork ... | 7a3ffef70339c10f8d65a27b88cd73bfbe13eb8a | 1e0d12b6f273d7345e32c16cd937475ed7c512ad | 19 | medium | [
{
"filename": "packages/react-native-renderer/src/ReactNativeTypes.js",
"patch": "@@ -34,7 +34,7 @@ export type AttributeType<T, V> =\n export type AnyAttributeType = AttributeType<$FlowFixMe, $FlowFixMe>;\n \n export type AttributeConfiguration = $ReadOnly<{\n- [propName: string]: AnyAttributeType,\n+ [p... |
vuejs/vue | 9,598 | fix(compiler): set end location for incomplete elements | <!--
Please make sure to read the Pull Request Guidelines:
https://github.com/vuejs/vue/blob/dev/.github/CONTRIBUTING.md#pull-request-guidelines
-->
<!-- PULL REQUEST TEMPLATE -->
<!-- (Update "[ ]" to "[x]" to check a box) -->
**What kind of change does this PR introduce?** (check at least one)
- [x] Bugf... | 7912f75c5eb09e0aef3e4bfd8a3bb78cad7540d7 | cbad54aa52847cfc934bb925d53c53ee57fc153d | 9 | medium | [
{
"filename": "src/compiler/parser/html-parser.js",
"patch": "@@ -279,7 +279,7 @@ export function parseHTML (html, options) {\n ) {\n options.warn(\n `tag <${stack[i].tag}> has no matching end tag.`,\n- { start: stack[i].start }\n+ { start: stack[i].start,... |
nodejs/node | 61,336 | src: fix pointer alignment | The NgLibMemoryManager::ReallocImpl method prefixes the allocated memory with its size, and returns a pointer to the region after it. This pointer can however no longer be suitably aligned.
On Arm 32bits this resulted in unaligned accesses, since the NEON vst1.64 instruction was used with a not properly aligned addr... | null | e74b42842e768489423b7bde0767109327dc4f54 | null | low | [
{
"filename": "src/node_mem-inl.h",
"patch": "@@ -8,6 +8,8 @@\n \n namespace node {\n namespace mem {\n+static constexpr size_t kReserveSizeAndAlign =\n+ std::max(sizeof(size_t), alignof(max_align_t));\n \n template <typename Class, typename AllocatorStruct>\n AllocatorStruct NgLibMemoryManager<Class, Al... |
facebook/react | 33,670 | Special case printing Promises in Performance Track Properties | Before:
<img width="266" alt="Screenshot 2025-06-30 at 8 32 23 AM" src="https://github.com/user-attachments/assets/98aae5e1-4b2c-49bd-9b71-040b788c36ba" />
After:
<img width="342" alt="Screenshot 2025-06-30 at 8 39 17 AM" src="https://github.com/user-attachments/assets/cd91c4a6-f6ae-4bec-9cd9-f42f4af468fe" />
| d2a288febf61a1755b78ce98b3cb17dd412b81e3 | e9cab42ece435ac3478ec85847e352177e596ae0 | 7 | medium | [
{
"filename": "packages/shared/ReactPerformanceTrackProperties.js",
"patch": "@@ -145,6 +145,40 @@ export function addValueToProperties(\n return;\n }\n }\n+ if (objectName === 'Promise') {\n+ if (value.status === 'fulfilled') {\n+ // Print the inner ... |
facebook/react | 33,664 | [Flight] Resolve Deep Cycles | Stacked on #33666.
If we ever get a future reference to a cycle and that reference gets eagerly parsed before the target has loaded then we can end up with a cycle that never gets resolved. That's because our cycle resolution only works if the cyclic future reference is created synchronously within the parsing path ... | 9c2a8dd5f87266308a8a8e55fb7f029ba530d36a | 3cfcdfb30720a5b8de0e981c8fdabec1abb61588 | 1 | medium | [
{
"filename": "packages/react-client/src/ReactFlightClient.js",
"patch": "@@ -165,16 +165,16 @@ const HALTED = 'halted'; // DEV-only. Means it never resolves even if connection\n \n type PendingChunk<T> = {\n status: 'pending',\n- value: null | Array<(T) => mixed>,\n- reason: null | Array<(mixed) => mix... |
ggml-org/llama.cpp | 20,552 | fix: prevent nullptr dereference | When encountering an unsupported template (e.g. translategemma), the code currently dereferences a nullptr and causes the program to crash.
With this fix, a proper exception will be thrown from `common_chat_templates_apply_jinja` instead. | null | 559646472d4b0ab732b977fefeebc3c38eae458b | null | low | [
{
"filename": "common/chat-diff-analyzer.cpp",
"patch": "@@ -479,6 +479,7 @@ analyze_content::analyze_content(const common_chat_template & tmpl, const analyz\n \n if (!comparison_with_tools || !comparison_with_reasoning) {\n LOG_DBG(ANSI_ORANGE \"%s: Template application failed\\n\" ANSI_RESET, ... |
vercel/next.js | 88,939 | Add Graphite workflow Cursor command | ### What?
Adds a Cursor command file (`.cursor/commands/gt-workflow.md`) that provides guidance on using Graphite (gt) instead of git for branch and commit operations.
### Why?
To make the Graphite workflow easily accessible within Cursor IDE via the `/gt-workflow` command, helping contributors follow the repository... | null | 631b2c60c89352e794b7d863db79457a278fe2ac | null | low | [
{
"filename": ".cursor/commands/gt-workflow.md",
"patch": "@@ -0,0 +1,67 @@\n+# Git Workflow with Graphite\n+\n+## Overview\n+\n+Use Graphite (`gt`) instead of git for ALL branch and commit operations in this repository.\n+\n+## Forbidden Git Commands\n+\n+NEVER use these git commands directly:\n+\n+- `git ... |
facebook/react | 33,666 | [Flight] Ensure we dedupe references if we later discover that it's the model root | I noticed we weren't deduping these cases. | bfc8801e0f0bfacb46bc71244e8244736bd995f4 | 9c2a8dd5f87266308a8a8e55fb7f029ba530d36a | 4 | medium | [
{
"filename": "packages/react-server-dom-webpack/src/__tests__/ReactFlightDOMBrowser-test.js",
"patch": "@@ -2574,4 +2574,54 @@ describe('ReactFlightDOMBrowser', () => {\n expect(errors).toEqual([new Error('Connection closed.')]);\n expect(container.innerHTML).toBe('');\n });\n+\n+ it('can dedupe... |
ollama/ollama | 8,643 | benchmark: performance of running ollama server | This PR introduces a benchmarking framework for measuring Ollama's inference performance across different models and scenarios. The implementation measures Time To First Token (TTFT), total generation time, and tokens per second throughput.
## Key Features
- Measures both cold start and warm start performance
- T... | null | fb6252d786c9cafdd2dbfa434535d85c611c0ff0 | null | low | [
{
"filename": "benchmark/server_benchmark_test.go",
"patch": "@@ -0,0 +1,178 @@\n+package benchmark\n+\n+import (\n+\t\"context\"\n+\t\"flag\"\n+\t\"fmt\"\n+\t\"testing\"\n+\t\"time\"\n+\n+\t\"github.com/ollama/ollama/api\"\n+)\n+\n+// Command line flags\n+var modelFlag string\n+\n+func init() {\n+\tflag.St... |
electron/electron | 49,499 | fix: issues with xdg activation on Linux | #### Description of Change
Refs CL:6147736
Refs CL:7122102
Refs CL:6931376
Fixes an issue where XDG Activation doesn't work when window is already open. See [this comment](https://issuetracker.google.com/issues/385007283#comment2) for further explanation
From CL:
> FileManager1 API was preferred over Open... | d689f76ff965d239decc5936c44494d8f15fea50 | e033c1007562eb3b79ef827a7e797848dfe93e10 | 30 | medium | [
{
"filename": "shell/common/platform_util_linux.cc",
"patch": "@@ -7,7 +7,9 @@\n #include <fcntl.h>\n \n #include <stdio.h>\n+#include <memory>\n #include <optional>\n+#include <queue>\n #include <string>\n #include <vector>\n \n@@ -27,8 +29,13 @@\n #include \"base/run_loop.h\"\n #include \"base/strings/esc... |
huggingface/transformers | 43,219 | Fix tokenizer auto_map being ignored for custom models | Fixes #43202
PR #42894 introduced an early-exit to `TokenizersBackend` when `tokenizer_class` doesn't match the registered tokenizer for a `model_type`. However, this check was placed before the `auto_map` extraction, causing custom tokenizers (with `trust_remote_code=True`) to be ignored.
Reproduction:
```python
fro... | null | 541046564da80b8ff655e1deb6c7ac15a6eed23a | null | low | [
{
"filename": "src/transformers/models/auto/tokenization_auto.py",
"patch": "@@ -625,9 +625,21 @@ def from_pretrained(\n # Next, let's try to use the tokenizer_config file to get the tokenizer class.\n tokenizer_config = get_tokenizer_config(pretrained_model_name_or_path, **kwargs)\n ... |
vuejs/vue | 9,550 | test: Add a unit test for "$off event and this event added by $once" | <!--
Please make sure to read the Pull Request Guidelines:
https://github.com/vuejs/vue/blob/dev/.github/CONTRIBUTING.md#pull-request-guidelines
-->
<!-- PULL REQUEST TEMPLATE -->
<!-- (Update "[ ]" to "[x]" to check a box) -->
**What kind of change does this PR introduce?** (check at least one)
- [ ] Bugf... | null | 1574ade607499bf4e3cf270c69249893e29eb259 | null | low | [
{
"filename": "test/unit/features/instance/methods-events.spec.js",
"patch": "@@ -56,6 +56,13 @@ describe('Instance methods events', () => {\n expect(spy).toHaveBeenCalledWith(1, 2, 3)\n })\n \n+ it('$off event added by $once', () => {\n+ vm.$once('test', spy)\n+ vm.$off('test', spy) // test of... |
facebook/react | 33,656 | [Flight] Don't replay performance logs when replayConsoleLogs is false | This is the same principle. They're both side-effects and go to the `console.*` namespace.
| bfc8801e0f0bfacb46bc71244e8244736bd995f4 | 811e203ed42c1a496790426a687d5045c473653d | 3 | medium | [
{
"filename": "packages/react-client/src/ReactFlightClient.js",
"patch": "@@ -702,14 +702,16 @@ export function reportGlobalError(response: Response, error: Error): void {\n }\n }\n if (enableProfilerTimer && enableComponentPerformanceTrack) {\n- markAllTracksInOrder();\n- flushComponentPerfor... |
nodejs/node | 61,344 | child_process: treat ipc length header as unsigned uint32 | Fixes: https://github.com/nodejs/node/issues/61312
<!--
Before submitting a pull request, please read:
- the CONTRIBUTING guide at https://github.com/nodejs/node/blob/HEAD/CONTRIBUTING.md
- the commit message formatting guidelines at
https://github.com/nodejs/node/blob/HEAD/doc/contributing/pull-requests.md#... | null | 5f982e28793ed001f5adb8fabea1d2a1f3247c71 | null | low | [
{
"filename": "lib/internal/child_process/serialization.js",
"patch": "@@ -74,12 +74,12 @@ const advanced = {\n while (messageBufferHead.length >= 4) {\n // We call `readUInt32BE` manually here, because this is faster than first converting\n // it to a buffer and using `readUInt32BE` on that... |
ollama/ollama | 9,923 | server/internal/client/ollama: persist through chunk download errors | This changes the behavior of the registry client to persist through
chunk download errors. The client will no longer bring down all
in-flight chunk downloads if one of them fails. Instead, it will
continue to download the other chunks, and once all of them are
complete, it will return an error if any of them failed... | 2ddacd7516aa78d2ad4899681c7a7f9af2eedc65 | c794fef2f27d141393064665d2774b341d091393 | 14 | medium | [
{
"filename": "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 ... |
huggingface/transformers | 43,405 | Fix failing `recurrent_gemma` tests | # What does this PR do?
Fixes these failing [recurrent_gemma](https://github.com/huggingface/transformers/actions/runs/21199210283/job/60982514314#step:14:6060) tests.
<img width="2311" height="472" alt="image" src="https://github.com/user-attachments/assets/75b3122d-714a-4e95-8a99-de3886358ceb" />
#... | eff263cd595bb47013af292ba619447cc842f4ab | 1f1532adcaa8cf873ca6586ff31a4120f544b751 | 7 | medium | [
{
"filename": "src/transformers/models/recurrent_gemma/modeling_recurrent_gemma.py",
"patch": "@@ -219,7 +219,7 @@ def forward(\n # Partial rotary embedding\n query_rot, query_pass = torch.chunk(query_states, int(1 / self.partial_rotary_factor), dim=-1)\n key_rot, key_pass = torch.ch... |
ggml-org/llama.cpp | 20,526 | codeowners : use teams | Use the new teams instead of individual users where applicable.
Is @am17an joining @ggml-org/ggml-cuda team?
Should server and webui be split individually as today, or add both to @ggml-org/server team? | 9789c4ecdc01d571331c14e5197514b53839de4b | cf45437d35eecd3f05a21968a4958a50e8038fb6 | 21 | medium | [
{
"filename": "CODEOWNERS",
"patch": "@@ -2,29 +2,13 @@\n # multiplie collaborators per item can be specified\n \n /.devops/*.Dockerfile @ngxson\n-/.github/actions/ @CISC\n-/.github/workflows/ @CISC\n+/.github/actions/ @ggml-o... |
electron/electron | 49,621 | fix: handle out of order recording errors in skia graphite | Backport of #49608
See that PR for details.
Notes: Fixed rendering glitches on macOS.
| 86196dc5886b4023141b2e12d408dc1367ef5fb9 | 594b38fb7d1a57ad893d871ff9c99edcfcbfd88f | 16 | medium | [
{
"filename": "patches/chromium/.patches",
"patch": "@@ -143,3 +143,4 @@ fix_check_for_file_existence_before_setting_mtime.patch\n fix_linux_tray_id.patch\n expose_gtk_ui_platform_field.patch\n fix_os_crypt_async_cookie_encryption.patch\n+graphite_handle_out_of_order_recording_errors.patch",
"additions"... |
ggml-org/llama.cpp | 20,585 | ci : make cross-build workflows manual | cont #20546
- Run cross-platform builds only via manual triggers
- Run on loongarch and spacemit code changes
- Run once every week | e445bb6f202bda60401d06185722951234e945a2 | 4057c5987d329772e530a9802e0f7659814323a5 | 1 | medium | [
{
"filename": ".github/workflows/build-cache.yml",
"patch": "@@ -37,31 +37,31 @@ jobs:\n path: ./vulkan_sdk\n version: ${{ env.VULKAN_SDK_VERSION }}\n \n- ubuntu-24-spacemit-cache:\n- runs-on: ubuntu-24.04\n-\n- env:\n- # Make sure this is in sync with build-linux-cross.yml\n... |
electron/electron | 49,613 | fix: fix Windows MSIX release build errors | #### Description of Change
Follow up to https://github.com/electron/electron/pull/49230
This PR fixes an error on release builds that seems to be caused by the new MSIX logic.
#### Checklist
<!-- Remove items that do not apply. For completed items, change [ ] to [x]. -->
- [ ] PR description included
- [ ... | 331d1e16f5b20cb453741af46842448383663a60 | 4b5d5f9dd5bb7eadec299060fc73ba1178388feb | 3 | medium | [
{
"filename": "BUILD.gn",
"patch": "@@ -420,6 +420,37 @@ action(\"electron_generate_node_defines\") {\n args = [ rebase_path(target_gen_dir) ] + rebase_path(inputs)\n }\n \n+# MSIX updater needs to be in a separate source_set because it uses C++/WinRT\n+# headers that require exceptions to be enabled.\n+s... |
ollama/ollama | 9,917 | Revert "parser: remove role validation from Modelfile parser (#9874)" | This reverts commit ffbfe833da387f9b6806fe887b85992c11d26eaa. | ffbfe833da387f9b6806fe887b85992c11d26eaa | 00ebda8cc4f2031096973b26d6b0de7425a9ca82 | 9 | medium | [
{
"filename": "parser/parser.go",
"patch": "@@ -7,7 +7,6 @@ import (\n \t\"errors\"\n \t\"fmt\"\n \t\"io\"\n-\t\"log/slog\"\n \t\"net/http\"\n \t\"os\"\n \t\"os/user\"\n@@ -301,8 +300,9 @@ const (\n )\n \n var (\n-\terrMissingFrom = errors.New(\"no FROM line\")\n-\terrInvalidCommand = errors.New(\"comman... |
ggml-org/llama.cpp | 20,588 | ci : make msys workflows manual | cont https://github.com/ggml-org/llama.cpp/pull/20546
- Run msys2 builds only via manual triggers
- Run once per week
| 87c5651ec7ed4d9a1736e2025b178c99cbe552e2 | e445bb6f202bda60401d06185722951234e945a2 | 1 | medium | [
{
"filename": ".github/workflows/build-msys.yml",
"patch": "@@ -0,0 +1,72 @@\n+name: CI (msys)\n+\n+on:\n+ # only manual triggers due to low-importance of the workflows\n+ # TODO: for regular runs, provision dedicated self-hosted runners\n+ workflow_dispatch:\n+ # run once every week\n+ schedule:\n+ ... |
ollama/ollama | 9,892 | 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 - since the fully context size is allocated u... | 3ed7ad3ab32b458aa2fdb8d0144c546efdb26a72 | 2d6eac9084a29060ccff69014e28e206a3a7a663 | 1 | medium | [
{
"filename": "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 <... |
vercel/next.js | 88,801 | Turbopack: query conditions in rules follow-ups | Rebase of https://github.com/vercel/next.js/pull/87291 | null | 98fd3456fb19b1305816ccba8f45d43eb1e367da | null | low | [
{
"filename": "crates/next-core/src/next_config.rs",
"patch": "@@ -2106,6 +2106,13 @@ mod tests {\n \"browser\",\n {\n \"path\": { \"type\": \"glob\", \"value\": \"*.svg\"},\n+ \"query\": {\n+ ... |
vuejs/vue | 9,572 | fix #9571, rendering async components after initial context was destroyed | <!--
Please make sure to read the Pull Request Guidelines:
https://github.com/vuejs/vue/blob/dev/.github/CONTRIBUTING.md#pull-request-guidelines
-->
<!-- PULL REQUEST TEMPLATE -->
<!-- (Update "[ ]" to "[x]" to check a box) -->
**What kind of change does this PR introduce?** (check at least one)
- [x] Bugf... | null | ed341137b23315b76ba391db1b0e537950c091e1 | null | low | [
{
"filename": "src/core/vdom/helpers/resolve-async-component.js",
"patch": "@@ -8,7 +8,8 @@ import {\n isTrue,\n isObject,\n hasSymbol,\n- isPromise\n+ isPromise,\n+ remove\n } from 'core/util/index'\n \n import { createEmptyVNode } from 'core/vdom/vnode'\n@@ -51,17 +52,21 @@ export function resolv... |
vuejs/vue | 9,563 | fix(types): allow scoped slots to return a single VNode | <!--
Please make sure to read the Pull Request Guidelines:
https://github.com/vuejs/vue/blob/dev/.github/CONTRIBUTING.md#pull-request-guidelines
-->
<!-- PULL REQUEST TEMPLATE -->
<!-- (Update "[ ]" to "[x]" to check a box) -->
**What kind of change does this PR introduce?** (check at least one)
- [x] Bugf... | f33301619d18b9392597c5230af17921c0b42466 | 241eea19a64550bfdb3f9d7e4197127997572842 | 1 | medium | [
{
"filename": "types/test/options-test.ts",
"patch": "@@ -323,6 +323,10 @@ Vue.component('component-with-scoped-slot', {\n item: (props: ScopedSlotProps) => [h('span', [props.msg])]\n }\n }),\n+ h('child', [\n+ // return single VNode (will be normalized to an array)\n+ ... |
facebook/react | 33,652 | devtools: emit performance entries only when profiling | ## Summary
This floods Timings track in dev mode and also hurts performance in dev.
Making sure we are buffering Performance entries (all of them are marks) only when profiling in RDT. This should be removed once we roll out Perf tracks.
| null | 58ac15cdc953a21ac1b352b6300abc3da87fa3d6 | null | low | [
{
"filename": "packages/react-devtools-shared/src/backend/profilingHooks.js",
"patch": "@@ -329,21 +329,17 @@ export function createProfilingHooks({\n }\n \n function markComponentRenderStarted(fiber: Fiber): void {\n- if (isProfiling || supportsUserTimingV3) {\n+ if (isProfiling) {\n const ... |
electron/electron | 49,596 | fix: duplicate fullscreen macOS menu item | Backport of #49074
See that PR for details.
Notes: Fixed an issue whereby a duplicate "Toggle Full Screen" menu item appeared in the View menu on macOS.
| 269a5393c09f7ab8c9ad28d09ef4402c54f4a931 | 356bba80607af85a49d95dc8c51154e2d409e765 | 16 | medium | [
{
"filename": "shell/browser/mac/electron_application_delegate.mm",
"patch": "@@ -56,11 +56,6 @@ - (void)willPowerOff:(NSNotification*)notify {\n }\n \n - (void)applicationWillFinishLaunching:(NSNotification*)notify {\n- // Don't add the \"Enter Full Screen\" menu item automatically.\n- [[NSUserDefaults s... |
nodejs/node | 60,946 | fs: fix errorOnExist behavior for directory copy in fs.cp | Fixes https://github.com/nodejs/node/issues/58947. This implements the errorOnExists check for onDir, which we previously had only for onFile.
<!--
Before submitting a pull request, please read:
- the CONTRIBUTING guide at https://github.com/nodejs/node/blob/HEAD/CONTRIBUTING.md
- the commit message formatting gu... | null | 24b16509d414cb2f48bb933109c4ac9c7303e0a9 | null | low | [
{
"filename": "lib/internal/fs/cp/cp.js",
"patch": "@@ -300,8 +300,17 @@ async function setDestTimestamps(src, dest) {\n return utimes(dest, updatedSrcStat.atime, updatedSrcStat.mtime);\n }\n \n-function onDir(srcStat, destStat, src, dest, opts) {\n+async function onDir(srcStat, destStat, src, dest, opts)... |
facebook/react | 33,654 | [Flight] Write Debug Info to Separate Priority Queue | This writes all debug info to a separate priority queue. In the future I'll put this on a different channel.
Ideally I think we'd put it in the bottom of the stream but because it actually blocks the elements from resolving anyway it ends up being better to put them ahead. At least for now.
When we have two separ... | d2a288febf61a1755b78ce98b3cb17dd412b81e3 | bfc8801e0f0bfacb46bc71244e8244736bd995f4 | 1 | medium | [
{
"filename": "packages/react-client/src/ReactFlightClient.js",
"patch": "@@ -764,6 +764,78 @@ function getTaskName(type: mixed): string {\n }\n }\n \n+function initializeElement(response: Response, element: any): void {\n+ if (!__DEV__) {\n+ return;\n+ }\n+ const stack = element._debugStack;\n+ co... |
vercel/next.js | 88,925 | [turbopack] add task type infromation to the print_cache_item_size feature | Enhance our debugging output to include information about TaskType values
example output:
```
Task type cache stats: 337.82MiB (364.22MiB)
49.98MiB (53.43MiB) <ModuleAssetContext as AssetContext>::resolve_asset x 417238 avg 125B (134B)
29.38MiB (31.46MiB) resolve x 297288 avg 103B (110B)
17.97MiB (22.58MiB) As... | c582bf8eb2e1fabb020c940f360562b70a26c6c2 | 20be9bb47a69e850ec474e0a60ea9e6d8e998d8a | 9 | medium | [
{
"filename": "turbopack/crates/turbo-tasks-backend/src/backend/mod.rs",
"patch": "@@ -1177,14 +1177,28 @@ impl<B: BackingStorage> TurboTasksBackendInner<B> {\n .into_iter()\n .collect::<Vec<_>>();\n if !task_cache_stats.is_empty() {\n+ ... |
huggingface/transformers | 43,401 | Fix failing `ChameleonIntegrationTests` | # What does this PR do?
Fixes these failing [ChameleonIntegrationTests](https://github.com/huggingface/transformers/actions/runs/21199210283/job/60982512511#step:14:5322)
<img width="2294" height="322" alt="Screenshot from 2026-01-22 07-24-39" src="https://github.com/user-attachments/assets/b3650c09-6a9c-4280-adb... | null | 1908fb6132126cae909a559685a78f3ebfe39356 | null | low | [
{
"filename": "tests/models/chameleon/test_modeling_chameleon.py",
"patch": "@@ -338,7 +338,7 @@ def test_model_7b(self):\n )\n prompt = \"<image>Describe what do you see here and tell me about the history behind it?\"\n \n- inputs = processor(images=image, text=prompt, return_tensors... |
electron/electron | 49,074 | fix: duplicate fullscreen macOS menu item | #### Description of Change
Closes https://github.com/electron/electron/issues/49048
This fixes an issue introduced in 38be633 whereby a duplicate "Toggle Full Screen" menu item appeared in the View menu on macOS.
When `menu.delegate` is set on an `NSMenu` that contains a full screen menu item, macOS automatica... | 64872a6cb8572f15af8a6ce43a6b6cb852195ec7 | 6825a522e9a535f421e102b4fce8463518429fd6 | 2 | medium | [
{
"filename": "shell/browser/mac/electron_application_delegate.mm",
"patch": "@@ -56,11 +56,6 @@ - (void)willPowerOff:(NSNotification*)notify {\n }\n \n - (void)applicationWillFinishLaunching:(NSNotification*)notify {\n- // Don't add the \"Enter Full Screen\" menu item automatically.\n- [[NSUserDefaults s... |
facebook/react | 33,655 | Include Component Props in Performance Track | Similar to how we can include a Promise resolved value we can include Component Props.
For now I left out props for Client Components for perf unless they error. I'll try it for Client Components in general in a separate PR.
<img width="730" alt="Screenshot 2025-06-26 at 5 54 29 PM" src="https://github.com/user-a... | bb6c9d521e7f44eb9a9754a14cde62c8e6439e5e | d2a288febf61a1755b78ce98b3cb17dd412b81e3 | 5 | medium | [
{
"filename": "packages/react-client/src/ReactFlightClient.js",
"patch": "@@ -100,7 +100,7 @@ import {getOwnerStackByComponentInfoInDev} from 'shared/ReactComponentInfoStack'\n \n import {injectInternals} from './ReactFlightClientDevToolsHook';\n \n-import {OMITTED_PROP_ERROR} from './ReactFlightPropertyAcc... |
ollama/ollama | 9,928 | fix: show correct bool value for kv in verbose show information | 2c8b4846437747bd23e7a176f83011e39ec2128b | 6d1103048eac63f27148d6d8fe47c98cbb6f184f | 22 | medium | [
{
"filename": "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... | |
ggml-org/llama.cpp | 20,489 | convert : support contiguous method on lora tensors | Fixes #20450 | null | 89d0aec0420f45718bc7f960134e7237dd4c9712 | null | low | [
{
"filename": "convert_lora_to_gguf.py",
"patch": "@@ -128,6 +128,12 @@ def size(self, dim=None):\n assert dim is None\n return self.shape\n \n+ def contiguous(self) -> LoraTorchTensor:\n+ return LoraTorchTensor(\n+ self._lora_A.contiguous(),\n+ self._lora_B.c... |
ollama/ollama | 9,912 | ollamarunner: Provide mechanism for backends to loading progress | This enables the runner to report progress back to the Ollama server, both for showing status to the user and also to prevent the server from killing the runner if it thinks things have stalled.
Most of the infrastructure was already there, this extends it to be available to the backends. | 0fbfcf3c9c7bfdbf4616238595eafd7eca2a916c | 0ff28758b3a5e9dc0149a93d87677dd0585590c1 | 2 | medium | [
{
"filename": "ml/backend.go",
"patch": "@@ -60,6 +60,10 @@ type CacheConfig struct {\n \n // BackendParams controls how the backend loads and executes models\n type BackendParams struct {\n+\t// Progress is a callback function that allows reporting percentage completion\n+\t// of model loading\n+\tProgress... |
electron/electron | 49,575 | build(dev-deps): bump @electron/lint-roller to 3.2.0 | Backport of #49565
See that PR for details.
Notes: none <!-- Please add a one-line description for app developers to read in the release notes, or 'none' if no notes relevant to app developers. Examples and help on special cases: https://github.com/electron/clerk/blob/main/README.md#examples -->
| cd00dcbafdceafc18a9b2c4881f9baea488a7fda | 0e901ca04c6961a4b20dfc683473a24403bf3874 | 11 | medium | [
{
"filename": "package.json",
"patch": "@@ -10,7 +10,7 @@\n \"@electron/docs-parser\": \"^2.0.0\",\n \"@electron/fiddle-core\": \"^1.3.4\",\n \"@electron/github-app-auth\": \"^3.2.0\",\n- \"@electron/lint-roller\": \"^3.1.2\",\n+ \"@electron/lint-roller\": \"^3.2.0\",\n \"@electron/typ... |
ollama/ollama | 9,918 | 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 original cache tensors as well.
Fixes #9904 | 0c220935bd9e23339af1c8f943f5418cbe1b705b | d3e9ca3eda5585204358f4ba23b3c7cb23bdb3e2 | 2 | high | [
{
"filename": "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... |
facebook/react | 33,649 | [Fizz] Rename ReactFizzContext to ReactFizzLegacyContext | #33622 forgot these. | cee7939b0017ff58230e19663c22393bfd9025ef | 31d91651e042e4939021f21a3d8799c13684a84b | 7 | medium | [
{
"filename": "packages/react-server/src/ReactFizzClassComponent.js",
"patch": "@@ -7,7 +7,7 @@\n * @flow\n */\n \n-import {emptyContextObject} from './ReactFizzContext';\n+import {emptyContextObject} from './ReactFizzLegacyContext';\n import {readContext} from './ReactFizzNewContext';\n \n import {disabl... |
vuejs/vue | 9,541 | fix(types): update this for nextTick api | **What kind of change does this PR introduce?** (check at least one)
- [x] Bugfix
- [ ] Feature
- [ ] Code style update
- [ ] Refactor
- [ ] Build-related changes
- [ ] Other, please describe:
**Does this PR introduce a breaking change?** (check one)
- [ ] Yes
- [x] No
If yes, please describe the impa... | null | f33301619d18b9392597c5230af17921c0b42466 | null | low | [
{
"filename": "types/test/vue-test.ts",
"patch": "@@ -86,6 +86,9 @@ class Test extends Vue {\n }\n });\n this.nextTick(() => {});\n+ this.nextTick(function () {\n+ console.log(this.text === 'test');\n+ }, { text: 'test'});\n this.nextTick().then(() => {});\n this.set({}, \"\... |
electron/electron | 49,565 | build(dev-deps): bump @electron/lint-roller to 3.2.0 | #### Description of Change
<!--
Thank you for your Pull Request. Please provide a description above and review
the requirements below.
Contributors guide: https://github.com/electron/electron/blob/main/CONTRIBUTING.md
-->
Picks up a minor feature which will let us switch some docs links that are currently h... | a7de47084bbd03e921df58a2e50ac01e72fb597a | 3d76d3a469d6fb22f0b6e137e210c81ae9e739fa | 3 | medium | [
{
"filename": "package.json",
"patch": "@@ -10,7 +10,7 @@\n \"@electron/docs-parser\": \"^2.0.0\",\n \"@electron/fiddle-core\": \"^1.3.4\",\n \"@electron/github-app-auth\": \"^3.2.0\",\n- \"@electron/lint-roller\": \"^3.1.2\",\n+ \"@electron/lint-roller\": \"^3.2.0\",\n \"@electron/typ... |
nodejs/node | 61,169 | test_runner: fix coverage report when a directory is named file | The coverage tree traversal checked `tree[key].file` to detect file
entries. When a directory named "file" contained a file also named
"file", this check incorrectly matched the child entry instead of
file metadata, causing a TypeError when accessing `.path`.
Check for `.file?.path` instead to correctly identify file ... | null | a00f5fea7029d3f0791ff6dddabc34b5a0ccf9b4 | null | low | [
{
"filename": "lib/internal/test_runner/utils.js",
"patch": "@@ -551,7 +551,7 @@ function getCoverageReport(pad, summary, symbol, color, table) {\n \n function printCoverageBodyTree(tree, depth = 0) {\n for (const key in tree) {\n- if (tree[key].file) {\n+ if (tree[key].file?.path) {\n ... |
facebook/react | 33,646 | [Flight] Emit start time before an await if one wasn't emitted already | There's a special case where if we create a new task, e.g. to serialize a promise like `<div>{promise}</div>` then that row doesn't have any start time emitted but it has a `task.time` inherited. We mostly don't need this because every other operation emits its own start time. E.g. when we started rendering a Server Co... | bb6c9d521e7f44eb9a9754a14cde62c8e6439e5e | 9406162bc95f1db78169a7f5e11bc56d324b2071 | 2 | high | [
{
"filename": "packages/react-server/src/ReactFlightServer.js",
"patch": "@@ -2211,6 +2211,10 @@ function emitAsyncSequence(\n debugInfo.stack = filterStackTrace(request, parseStackTrace(stack, 1));\n }\n }\n+ // We don't have a start time for this await but in case there was no start t... |
vercel/next.js | 88,930 | Update Rspack production test manifest | This auto-generated PR updates the production integration test manifest used when testing Rspack. | 4e8671cf5c142b1661b6b53404782aace2fb8d71 | 403538a6f8b409e9b62e89b2d25829e10cd68681 | 23 | medium | [
{
"filename": "test/rspack-build-tests-manifest.json",
"patch": "@@ -3556,6 +3556,17 @@\n \"flakey\": [],\n \"runtimeError\": false\n },\n+ \"test/e2e/app-dir/mdx-font-preload/mdx-font-preload.test.ts\": {\n+ \"passed\": [\n+ \"mdx-font-preload should apply font class from layout\",\n+ ... |
huggingface/transformers | 43,389 | Fix always issuing generation deprecation warning | # What does this PR do?
**Note:** already discussed with @zucchini-nlp internally
Commit 38e5987 from PR #43194 introduced a minor bug that causes the issuing of a deprecation warning regardless of the user using the deprecated interface.
For example:
```
model.generate(**input_ids, do_sample=False)
```
wo... | null | 6e1cde6780e40175e8729e4c095bda64601347f0 | null | low | [
{
"filename": "src/transformers/generation/utils.py",
"patch": "@@ -1812,7 +1812,7 @@ def _prepare_generation_config(\n generation_config.cache_implementation = None\n \n # It doesn't make sense to allow kwargs and `generation_config`, that should be mutually exclusive\n- if gener... |
facebook/react | 33,641 | [Flight] Log aborted await and component renders | <img width="926" alt="Screenshot 2025-06-25 at 1 02 14 PM" src="https://github.com/user-attachments/assets/1877d13d-5259-4cc4-8f48-12981e3073fe" />
The I/O entry doesn't show as aborted in the Server Request track because technically it wasn't. The end time is just made up. It's still going. It's not aborted until t... | cee7939b0017ff58230e19663c22393bfd9025ef | bb6c9d521e7f44eb9a9754a14cde62c8e6439e5e | 4 | medium | [
{
"filename": "packages/react-client/src/ReactFlightClient.js",
"patch": "@@ -77,10 +77,12 @@ import {\n markAllTracksInOrder,\n logComponentRender,\n logDedupedComponentRender,\n+ logComponentAborted,\n logComponentErrored,\n logIOInfo,\n logIOInfoErrored,\n logComponentAwait,\n+ logCompone... |
electron/electron | 49,595 | fix: duplicate fullscreen macOS menu item | Backport of #49074
See that PR for details.
Notes: Fixed an issue whereby a duplicate "Toggle Full Screen" menu item appeared in the View menu on macOS.
| 457598a261d196e81d6033692feef96757f84444 | f6df2854ff33d062709eecbc33de11d3c0bbff69 | 6 | medium | [
{
"filename": "shell/browser/mac/electron_application_delegate.mm",
"patch": "@@ -56,11 +56,6 @@ - (void)willPowerOff:(NSNotification*)notify {\n }\n \n - (void)applicationWillFinishLaunching:(NSNotification*)notify {\n- // Don't add the \"Enter Full Screen\" menu item automatically.\n- [[NSUserDefaults s... |
ggml-org/llama.cpp | 20,460 | Guard against sumq2 being 0 in IQ4_NL resulting in nan values | With `IQ4_NL` on several recent models there have been issues where during quantization NaN blocks are being found which crashes the quant
It seems to be stemming from a scenario where `sumq2` is 0 for a given block, likely from not having imatrix data for some obscure expert, or the weights themselves being 0 as we... | null | b9da4444dfded18730eaa2a64f4e377e274d49da | null | low | [
{
"filename": "ggml/src/ggml-quants.c",
"patch": "@@ -4767,7 +4767,7 @@ static void quantize_row_iq4_nl_impl(const int super_block_size, const int block\n sumqx += w*q*xb[j];\n sumq2 += w*q*q;\n }\n- d = sumqx/sumq2;\n+ d = sumq2 > 0 ? sumqx/sumq2 : 0.f;\n ... |
vuejs/vue | 9,595 | fix: avoid compression of unicode sequences by using regexps, closes #9456 | closes #9456
<!--
Please make sure to read the Pull Request Guidelines:
https://github.com/vuejs/vue/blob/dev/.github/CONTRIBUTING.md#pull-request-guidelines
-->
<!-- PULL REQUEST TEMPLATE -->
<!-- (Update "[ ]" to "[x]" to check a box) -->
**What kind of change does this PR introduce?** (check at least on... | null | 7912f75c5eb09e0aef3e4bfd8a3bb78cad7540d7 | null | low | [
{
"filename": "src/compiler/parser/html-parser.js",
"patch": "@@ -11,12 +11,12 @@\n \n import { makeMap, no } from 'shared/util'\n import { isNonPhrasingTag } from 'web/compiler/util'\n-import { unicodeLetters } from 'core/util/lang'\n+import { unicodeRegExp } from 'core/util/lang'\n \n // Regular Expressio... |
nodejs/node | 61,392 | test-runner: fix rerun ambiguous test failures | this is a fix for the previously added fature `--test-rerun-failures`
in a case there are nested ambiguous tests with the same name and same exact location, a following rerun won't run the failing test.
see the changes to the tests in this PR for a reproduction
the most effective way to solve this is to emit tests i... | 538aebfa5d7b28c9f569c0ea52fca10d15c8aa3c | 8365edcbd0d98fdd79993a61302e35d299c3d557 | 19 | medium | [
{
"filename": "lib/internal/test_runner/reporter/rerun.js",
"patch": "@@ -1,6 +1,7 @@\n 'use strict';\n \n const {\n+ ArrayPrototypeMap,\n ArrayPrototypePush,\n JSONStringify,\n } = primordials;\n@@ -11,19 +12,55 @@ function reportReruns(previousRuns, globalOptions) {\n return async function reporter... |
electron/electron | 49,598 | fix: duplicate fullscreen macOS menu item | Backport of #49074
See that PR for details.
Notes: Fixed an issue whereby a duplicate "Toggle Full Screen" menu item appeared in the View menu on macOS.
| 32281a6d08b4498d5bfad1e00465b5db3eb3f8bf | d7bdf928174c99cc674ed1f6bfc7e026e88f7e1f | 6 | medium | [
{
"filename": "shell/browser/mac/electron_application_delegate.mm",
"patch": "@@ -56,11 +56,6 @@ - (void)willPowerOff:(NSNotification*)notify {\n }\n \n - (void)applicationWillFinishLaunching:(NSNotification*)notify {\n- // Don't add the \"Enter Full Screen\" menu item automatically.\n- [[NSUserDefaults s... |
ollama/ollama | 9,874 | parser: remove role validation from Modelfile parser | This allows passing arbitrary roles in a MESSAGE block in a Modelfile. Related to https://github.com/ollama/ollama-python/pull/462 and comments from @ParthSareen.
Example Modelfiles that this PR will now allow:
```
FROM granite3.2
MESSAGE control thinking
```
or
```
FROM granite3.2
MESSAGE somerandom... | null | ffbfe833da387f9b6806fe887b85992c11d26eaa | null | low | [
{
"filename": "parser/parser.go",
"patch": "@@ -7,6 +7,7 @@ import (\n \t\"errors\"\n \t\"fmt\"\n \t\"io\"\n+\t\"log/slog\"\n \t\"net/http\"\n \t\"os\"\n \t\"os/user\"\n@@ -300,9 +301,8 @@ const (\n )\n \n var (\n-\terrMissingFrom = errors.New(\"no FROM line\")\n-\terrInvalidMessageRole = errors.New(... |
electron/electron | 49,588 | build(dev-deps): bump @electron/lint-roller to 3.2.0 | Backport of #49565.
See that PR for details.
Notes: none | 87badb84f8a9808b42f35f0806d83ddcefd5f484 | e1ac9d5d1bec5798d7464f07c74242b73ad88d56 | 6 | medium | [
{
"filename": "package.json",
"patch": "@@ -10,7 +10,7 @@\n \"@electron/docs-parser\": \"^2.0.0\",\n \"@electron/fiddle-core\": \"^1.3.4\",\n \"@electron/github-app-auth\": \"^3.2.0\",\n- \"@electron/lint-roller\": \"^3.1.2\",\n+ \"@electron/lint-roller\": \"^3.2.0\",\n \"@electron/typ... |
vercel/next.js | 88,931 | Update Rspack development test manifest | This auto-generated PR updates the development integration test manifest used when testing Rspack. | null | 1fbb35f2037c944823d5875a12fcbf717f7cc04b | null | low | [
{
"filename": "test/rspack-dev-tests-manifest.json",
"patch": "@@ -522,6 +522,33 @@\n \"flakey\": [],\n \"runtimeError\": false\n },\n+ \"test/development/app-dir/browser-log-forwarding/fixtures/error-level/error-level.test.ts\": {\n+ \"passed\": [\n+ \"browser-log-forwarding error level ... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.