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 |
|---|---|---|---|---|---|---|---|---|
ollama/ollama | 14,406 | ggml: ensure tensor size is valid | When quantizing tensors during model creation validate that the resulting sizes match what is expected based on the shape. | null | 9d902d63ce9e741c8c9f0b9716183905785e132e | null | low | [
{
"filename": "fs/ggml/gguf.go",
"patch": "@@ -245,7 +245,22 @@ func (llm *gguf) Decode(rs io.ReadSeeker) error {\n \tpadding := ggufPadding(offset, int64(alignment))\n \tllm.tensorOffset = uint64(offset + padding)\n \n+\t// get file size to validate tensor bounds\n+\tfileSize, err := rs.Seek(0, io.SeekEnd)... |
huggingface/transformers | 44,990 | More small vllm fixes | # What does this PR do?
- BC for check model inputs | null | 0e1978c9eb69ec64b55245212dbf63deab19d25b | null | low | [
{
"filename": "src/transformers/utils/generic.py",
"patch": "@@ -940,6 +940,14 @@ def wrapper(self, *args, **kwargs):\n return wrapper\n \n \n+# bc for check_model_inputs:\n+\n+\n+def check_model_inputs(func):\n+ logger.warning_once(\"The `check_model_inputs` decorator is deprecated in favor of `merg... |
vercel/next.js | 91,487 | Buffer prefetch response before passing to Flight client | **Current:**
1. https://github.com/vercel/next.js/pull/91487
**Up next:**
2. https://github.com/vercel/next.js/pull/91488
3. https://github.com/vercel/next.js/pull/89297
---
Prefetch responses include metadata (in the Flight stream sense, not HTML document metadata) that describes properties of the overall respons... | 5dd48e3598b1d70188e7af5828ea724aa3eb8203 | d50bef07caa60917681b0ced0f1bc523c2be4bba | 26 | medium | [
{
"filename": "packages/next/src/client/components/segment-cache/cache.ts",
"patch": "@@ -1682,7 +1682,13 @@ export async function fetchRouteOnCacheMiss(\n const couldBeIntercepted =\n varyHeader !== null && varyHeader.includes(NEXT_URL)\n \n- // Track when the network connection closes.\n+ ... |
nodejs/node | 62,471 | test: wait for reattach before initial break on restart | `restart` can return to the prompt before the debugger reattaches and emits the
initial break. That leaves a race in these tests, which still flakes on macOS
unusual-path runs.
This change waits for `Debugger attached.` and the prompt before
`waitForInitialBreak()`.
CI failed logs:
- test/parallel/test-debugg... | 9137925f60886c17e71f7e2e1bd33aed8eaee034 | d9645d75e59d392f0e1e2fa7efb116a231a40f22 | 27 | medium | [
{
"filename": "test/parallel/test-debugger-preserve-breaks.js",
"patch": "@@ -29,6 +29,8 @@ const script = path.relative(process.cwd(), scriptFullPath);\n await cli.stepCommand('c'); // hit line 3\n assert.deepStrictEqual(cli.breakInfo, { filename: script, line: 3 });\n await cli.command('restar... |
facebook/react | 35,691 | Fix exhaustive deps bug with flow type casting. | Summary:
I noticed there's a bug where the lint will recognize the type on a cast annotation as a missing dependency;
```
function MyComponent() {
type ColumnKey = 'id' | 'name';
type Item = {id: string, name: string};
const columns = useMemo(
() => [
... | null | c137dd6f5484f1bb68a64f07dd0e40d474a29759 | null | low | [
{
"filename": "packages/eslint-plugin-react-hooks/__tests__/ESLintRuleExhaustiveDeps-test.js",
"patch": "@@ -7913,6 +7913,25 @@ const testsFlow = {\n }\n `,\n },\n+ // Flow type aliases in type assertions should not be flagged as missing dependencies\n+ {\n+ code: normalizeInden... |
rust-lang/rust | 153,369 | library/test: always enable unstable features for miri | The unstable features of the `test` crate used to be default-enabled,
and manually disabled when building the beta and stable channels. Commit
dae8ea92a733 flipped that to default-disabled, only enabled for nightly
and dev channels.
However, this broke miri testing on the beta/stable channels, because it
also us... | null | 2feb90a5e3e589791df746ef37333ff956f1b4f6 | null | low | [
{
"filename": "library/test/build.rs",
"patch": "@@ -2,11 +2,16 @@ fn main() {\n println!(\"cargo:rerun-if-changed=build.rs\");\n println!(\"cargo:rustc-check-cfg=cfg(enable_unstable_features)\");\n \n- let rustc = std::env::var(\"RUSTC\").unwrap_or_else(|_| \"rustc\".into());\n- let version =... |
ollama/ollama | 14,217 | site: update readme | 379fd64fa837e51eddeda6c32f5c812d912a6751 | 519b11eba198576e1c6f2f60155420323e4edf1a | 1 | medium | [
{
"filename": "README.md",
"patch": "@@ -1,20 +1,30 @@\n-<div align=\"center\">\n- <a href=\"https://ollama.com\">\n- <img alt=\"ollama\" width=\"240\" src=\"https://github.com/ollama/ollama/assets/3325447/0d0b44e2-8f4a-4e99-9b52-a5c1c741c8f7\">\n+<p align=\"center\">\n+ <a href=\"https://ollama.com\">... | |
electron/electron | 50,587 | ci: add functionality for programmatic add/remove needs-signed-commits label (#50316) | #### Description of Change
Backport of https://github.com/electron/electron/pull/50316
#### Checklist
<!-- Remove items that do not apply. For completed items, change [ ] to [x]. -->
- [x] PR description included
- [ ] I have built and tested this PR
- [ ] `npm test` passes
- [ ] [PR release notes](https:/... | null | 5a69e80cacffbf8870124e61e593c8d3b40ff398 | null | low | [
{
"filename": ".github/workflows/build.yml",
"patch": "@@ -431,3 +431,30 @@ jobs:\n - name: GitHub Actions Jobs Done\n run: |\n echo \"All GitHub Actions Jobs are done\"\n+\n+ check-signed-commits:\n+ name: Check signed commits in green PR\n+ needs: gha-done\n+ if: ${{ contains(g... |
facebook/react | 35,689 | [compiler][snap] More minimization improvements | * A few new minimization strategies, removing function params and array/object pattern elements
* Ensure that we preserve the same set of errors based on not just category+reason but also description. | 3ce1316b05968d2a8cffe42a110f2726f2c44c3e | 90c6d1b218bbe07eeb757ce8777805a9b7f8e3c3 | 3 | medium | [
{
"filename": "compiler/packages/snap/src/minimize.ts",
"patch": "@@ -18,7 +18,7 @@ type CompileSuccess = {kind: 'success'};\n type CompileParseError = {kind: 'parse_error'; message: string};\n type CompileErrors = {\n kind: 'errors';\n- errors: Array<{category: string; reason: string}>;\n+ errors: Arra... |
facebook/react | 35,684 | [flags] Turn on `enableViewTransition` in OSS `react-test-renderer` | Just to get OSS builds in a consistent state. It's still dynamic in www and off in native-fb | e32c1261219a9d665f392a4799b8a24ea326f1a1 | f84ce5a45c47b1081a09c17eea58c16ef145c113 | 5 | medium | [
{
"filename": "packages/shared/forks/ReactFeatureFlags.test-renderer.js",
"patch": "@@ -60,7 +60,7 @@ export const enableEagerAlternateStateNodeCleanup: boolean = true;\n export const enableYieldingBeforePassive: boolean = true;\n \n export const enableThrottledScheduling: boolean = false;\n-export const en... |
huggingface/transformers | 44,899 | fix(models): Fix Perceiver interpolate_pos_encoding interpolating to the source size | ### What does this PR do?
The following failing Perceiver use case was identified and fixed in this PR:
→ c6d2848a23 ([🚨 Fix torch.jit.trace for interpolate_pos_encoding in all vision models](https://github.com/huggingface/transformers/pull/33226)) refactored all vision models' `interpolate_pos_encoding` methods... | null | 2620c4ddd41afcc8d3b1d1134803bd418ffa45a7 | null | low | [
{
"filename": "src/transformers/models/perceiver/modeling_perceiver.py",
"patch": "@@ -2567,7 +2567,7 @@ def interpolate_pos_encoding(self, position_embeddings: torch.Tensor, height: in\n \n position_embeddings = nn.functional.interpolate(\n position_embeddings,\n- size=(new_h... |
nodejs/node | 62,470 | test: disable flaky WPT Blob test on AIX | The test Blob-array-buffer.any.js can be flaky on AIX due to a recognized race condition in v8.
The failed assert is in deps/v8/src/base/platform/platform-aix.cc on line 199, preceeded by the comment:
https://github.com/nodejs/node/blob/5b6091ce31068e1c86f863273cdd397ecb47e3b7/deps/v8/src/base/platform/platform-a... | null | db887ea6c12889b8281002fb4b3a9a1fa9b28e00 | null | low | [
{
"filename": "test/wpt/status/FileAPI/blob.cjs",
"patch": "@@ -0,0 +1,59 @@\n+'use strict';\n+\n+const os = require('node:os');\n+\n+// On AIX, V8's OS::DecommitPages() has an inherent race condition caused by\n+// AIX's non-POSIX MAP_FIXED behavior. The implementation must munmap() then\n+// mmap(), and a... |
vercel/next.js | 91,509 | [react-sync] Allow locking in manual sync | Allows pointing the installed, vendored React at the currently vendored in source e.g.
```json
"react-server-dom-turbopack": "file:packages/next/src/compiled/react-server-dom-turbopack",
"react-server-dom-turbopack-experimental": "file:packages/next/src/compiled/react-server-dom-turbopack-experimental",
"react-ser... | null | c5577bb38eb41bee05c0de4a4b9f0bc53dd89fb0 | null | low | [
{
"filename": "packages/next/taskfile.js",
"patch": "@@ -1457,7 +1457,11 @@ export async function copy_vendor_react(task_) {\n // TODO-APP: remove unused fields from package.json and unused files\n function overridePackageName(source) {\n const json = JSON.parse(source)\n- json.name = jso... |
electron/electron | 50,586 | ci: add functionality for programmatic add/remove needs-signed-commits label (#50316) | #### Description of Change
Backport of https://github.com/electron/electron/pull/50316
#### Checklist
<!-- Remove items that do not apply. For completed items, change [ ] to [x]. -->
- [x] PR description included
- [ ] I have built and tested this PR
- [ ] `npm test` passes
- [ ] [PR release notes](https:/... | null | dc6d1265336b383b6db4e88a31fd21038d4f53c9 | null | low | [
{
"filename": ".github/workflows/build.yml",
"patch": "@@ -431,3 +431,30 @@ jobs:\n - name: GitHub Actions Jobs Done\n run: |\n echo \"All GitHub Actions Jobs are done\"\n+\n+ check-signed-commits:\n+ name: Check signed commits in green PR\n+ needs: gha-done\n+ if: ${{ contains(g... |
ollama/ollama | 14,376 | mlx: don't default to affine quantization for unquantized models | Otherwise the BF16 version of models trigger segfaults when they call into quantized kernels. | 041fb7763945d11e0484cb5c838b5fea67741c94 | f20dc6b698eba22c125b1b86a301d6e42673c85e | 26 | medium | [
{
"filename": "x/imagegen/safetensors/loader.go",
"patch": "@@ -37,9 +37,11 @@ func QuantizationParams(quantization string) (groupSize, bits int, mode string)\n \tcase \"MXFP8\":\n \t\t// Microsoft MX FP8: group_size=32, bits=8, E4M3 scales (no qbias)\n \t\treturn 32, 8, \"mxfp8\"\n-\tcase \"FP8\", \"Q8\", ... |
vercel/next.js | 91,488 | [Segment Cache] Always upsert on prefetch completion | **Previous:**
1. https://github.com/vercel/next.js/pull/91487
**Current:**
2. https://github.com/vercel/next.js/pull/91488
**Up next:**
3. https://github.com/vercel/next.js/pull/89297
---
When a prefetch response includes vary params, the segment cache rekeys entries to a more generic path based on which params ... | d50bef07caa60917681b0ced0f1bc523c2be4bba | e16f6fb2f333000737d5f1523a2f8f95240f4c98 | 1 | medium | [
{
"filename": "packages/next/src/client/components/segment-cache/cache.ts",
"patch": "@@ -1938,7 +1938,8 @@ export async function fetchSegmentOnCacheMiss(\n rejectSegmentCacheEntry(segmentCacheEntry, Date.now() + 10 * 1000)\n return null\n }\n- const staleAt = Date.now() + getStaleTimeMs(... |
rust-lang/rust | 154,511 | [perf] Revert FastISel patch | <!-- homu-ignore:start -->
*[View all comments](https://triagebot.infra.rust-lang.org/gh-comments/rust-lang/rust/pull/154511)*
<!-- homu-ignore:end -->
This caused a significant compile-time regression for debug builds.
There is another change (https://github.com/llvm/llvm-project/pull/186723) that mitigates that r... | null | 116458d0a5ae01cd517cabd2d1aee7f5457018ab | null | low | [
{
"filename": "src/llvm-project",
"patch": "@@ -1 +1 @@\n-Subproject commit 05918363362b439b9b0bced3daa14badd75da790\n+Subproject commit 1cb4e3833c1919c2e6fb579a23ac0e2b22587b7e",
"additions": 1,
"deletions": 1
}
] |
ollama/ollama | 14,379 | mlxrunner: Fix duplicate log prefixes and reduce log noise | Pass subprocess stdout/stderr through to the parent's stderr directly instead of re-wrapping each line with slog. The subprocess already writes structured slog output, so the re-wrapping produced nested timestamps, levels, and message fields that were hard to read.
Also downgrade verbose KV cache debug logs to trace... | 5c73c4e2eedeeb847743f43b4d2e85ff4b02be4b | 8daf47fb3afec28de3e447d974c2bf911bcafc20 | 2 | medium | [
{
"filename": "x/mlxrunner/cache/cache.go",
"patch": "@@ -3,8 +3,7 @@\n package cache\n \n import (\n-\t\"log/slog\"\n-\n+\t\"github.com/ollama/ollama/logutil\"\n \t\"github.com/ollama/ollama/x/mlxrunner/mlx\"\n )\n \n@@ -113,7 +112,7 @@ func (c *RotatingKVCache) Update(keys, values *mlx.Array) (*mlx.Array,... |
facebook/react | 35,648 | [Perf Tracks] Handle arrays with bigints in deep objects | Fixes https://github.com/facebook/react/issues/35004
Closes https://github.com/facebook/react/pull/35013
We show arrays with primitive values `JSON.stringify`ed in the performance tracks. However, `bigint` is not accepted by `JSON.stringify` so we have to treat arrays with these values as complex.
We could imple... | null | ff191f24b5f49252672ac4d3c46ef1e79cf7290d | null | low | [
{
"filename": "packages/react-reconciler/src/__tests__/ReactPerformanceTrack-test.js",
"patch": "@@ -440,4 +440,87 @@ describe('ReactPerformanceTracks', () => {\n ]);\n performanceMeasureCalls.length = 0;\n });\n+\n+ // @gate __DEV__ && enableComponentPerformanceTrack\n+ it('can handle bigint in... |
nodejs/node | 62,396 | src: improve EC JWK import performance | ```
./node benchmark/crypto/create-keyobject.js keyFormat=jwk-public n=1000000 keyType=ec
```
| Branch | ops/sec |
|---|---|
| main | 76,259 |
| improve-ec-jwk-import | 168,920 |
All EC JWK import supported curves go through the optimized branch but I left the old one in just in case. | 38647b388bbf5d350f8febe77a6674ec0d0607c5 | f3633ef4c8e4e3fcc27b293993d238e9fd3cfc46 | 21 | medium | [
{
"filename": "deps/ncrypto/ncrypto.cc",
"patch": "@@ -3538,8 +3538,38 @@ bool ECKeyPointer::setPublicKey(const ECPointPointer& pub) {\n bool ECKeyPointer::setPublicKeyRaw(const BignumPointer& x,\n const BignumPointer& y) {\n if (!key_) return false;\n- return EC_KEY_se... |
ollama/ollama | 14,336 | ui: use capability-based detection for web search | **Summary**
- Replace hardcoded model name prefix checks (gpt-oss, qwen3, deepseek-v3) with the existing CapabilityTools capability detection, so any model with tool support automatically gets web search
- Add useHasToolsCapability hook in the frontend (mirrors the existing useHasVisionCapability pattern)
- Remove... | fa6c0127e695916c1bf8b58650c232ddaf5a5012 | 6c980579cdc62d67aee13511102437a6fe5aa13a | 7 | medium | [
{
"filename": "app/ui/app/src/components/ChatForm.tsx",
"patch": "@@ -17,7 +17,10 @@ import {\n } from \"@/hooks/useChats\";\n import { useNavigate } from \"@tanstack/react-router\";\n import { useSelectedModel } from \"@/hooks/useSelectedModel\";\n-import { useHasVisionCapability } from \"@/hooks/useModelC... |
electron/electron | 50,585 | ci: add functionality for programmatic add/remove needs-signed-commits label (#50316) |
#### Description of Change
Backport of https://github.com/electron/electron/pull/50316
#### Checklist
<!-- Remove items that do not apply. For completed items, change [ ] to [x]. -->
- [x] PR description included
- [ ] I have built and tested this PR
- [ ] `npm test` passes
- [ ] [PR release notes](https... | 1b3e1433dd113c1c9a40e3247b45d9834ddc2415 | 45d03a53923d4fe0f28b5faf0e1785e8a96e3b59 | 21 | medium | [
{
"filename": ".github/workflows/build.yml",
"patch": "@@ -431,3 +431,30 @@ jobs:\n - name: GitHub Actions Jobs Done\n run: |\n echo \"All GitHub Actions Jobs are done\"\n+\n+ check-signed-commits:\n+ name: Check signed commits in green PR\n+ needs: gha-done\n+ if: ${{ contains(g... |
huggingface/transformers | 44,970 | Fix CPU 16 bytes alignment issue using equivalent fallback | # What does this PR do?
<!--
Congratulations! You've made it this far! You're not quite done yet though.
Once merged, your PR is going to appear in the release notes with the title you set, so make sure it's a great title that fully reflects the extent of your awesome contribution.
Then, please replace this w... | null | c3b41c6c57be42e5dc2834af25c0c15be98a6678 | null | low | [
{
"filename": "docs/source/en/weightconverter.md",
"patch": "@@ -168,7 +168,6 @@ Operations are fully reversible. Saving reverses the conversions and returns the\n | [`MergeModulelist(dim)`] | [`SplitModulelist(dim)`] |\n | [`SplitModulelist(dim)`] | [`MergeModulelist(dim)`] |\n | [`Transpose(d0, d1)`] | [`... |
ollama/ollama | 14,341 | Avoid Excessive MLX Memory Usage | This fixes two issues that caused unbounded memory use with MLX and eventual crashing:
- Replace error-prone reference counting for array lifecycles with a pin/sweep model — pin the arrays we need (outputs, cache), sweep everything else — reducing memory leaks while leveraging MLX's internal reference tracking
- Simp... | 5daf59cc6666dd036af8fab8c5df6b5571a9a9ba | 5c73c4e2eedeeb847743f43b4d2e85ff4b02be4b | 1 | medium | [
{
"filename": "x/mlxrunner/cache.go",
"patch": "@@ -3,94 +3,65 @@\n package mlxrunner\n \n import (\n+\t\"fmt\"\n \t\"log/slog\"\n \n+\t\"github.com/ollama/ollama/logutil\"\n \t\"github.com/ollama/ollama/x/mlxrunner/cache\"\n+\t\"github.com/ollama/ollama/x/mlxrunner/mlx\"\n )\n \n+// CacheEntry stores a sin... |
vercel/next.js | 91,528 | Skip catch-error-react-compiler test to unblock CI | https://github.com/vercel/next.js/actions/runs/23170738194/job/67327462789#step:36:318 is holding up CI, including release. Cause is not from `catchError` PR, it's just that the test has discovered missing support of Next.js for Pages + React 17/18 + React compiler. Will work in parallel. | fa32daad8d7d0cfff6de16d1fc13fe8bd2d1802e | 6db019c330d55eff6cf4ed7a093a0fa1c5a3daf3 | 28 | medium | [
{
"filename": "test/e2e/app-dir/catch-error/catch-error-react-compiler.test.ts",
"patch": "@@ -1,6 +1,12 @@\n import { nextTestSetup } from 'e2e-utils'\n \n-describe('app-dir - unstable_catchError with react compiler', () => {\n+// FIXME: If NEXT_TEST_REACT_VERSION is set, skip the test for now. Need to add... |
facebook/react | 35,679 | [Perf Tracks] Prevent crash when accessing `$$typeof` | Alternate to https://github.com/facebook/react/issues/34840
Fixes https://github.com/facebook/react/issues/34840
Closes https://github.com/facebook/react/issues/34840
## Summary
Guards dotting into `$$typeof` unconditionally. The try-catch proposed in https://github.com/facebook/react/issues/34840 feels too h... | da64117876002bd60baa9e93cad77421bbf0293f | 6853d7ab2f73989f31fb5a8ed652bd58225ead26 | 13 | medium | [
{
"filename": "packages/react-reconciler/src/__tests__/ReactPerformanceTrack-test.js",
"patch": "@@ -523,4 +523,110 @@ describe('ReactPerformanceTracks', () => {\n ],\n ]);\n });\n+\n+ // @gate __DEV__ && enableComponentPerformanceTrack\n+ it('diffs HTML-like objects', async () => {\n+ cons... |
rust-lang/rust | 154,344 | Update LLVM to 22.1.2 | Fixes https://github.com/rust-lang/rust/issues/154101. | null | d0fcb41318ef091cdd16e5157f7f5189a99bc359 | null | low | [
{
"filename": "src/llvm-project",
"patch": "@@ -1 +1 @@\n-Subproject commit 41f177ed26a5252a30ea6090a4da66ce0a96bf44\n+Subproject commit 05918363362b439b9b0bced3daa14badd75da790",
"additions": 1,
"deletions": 1
}
] |
ollama/ollama | 13,164 | app: open app instead of always navigating to / on connect | 1fd4cb87b202ddacbc45c2a14284acfafc552101 | 417a81fda3cefdb57e6b1c72963d016d238ab97d | 20 | medium | [
{
"filename": "app/cmd/app/app.go",
"patch": "@@ -397,8 +397,8 @@ func checkUserLoggedIn(uiServerPort int) bool {\n // handleConnectURLScheme fetches the connect URL and opens it in the browser\n func handleConnectURLScheme() {\n \tif checkUserLoggedIn(uiServerPort) {\n-\t\tslog.Info(\"user is already logge... | |
nodejs/node | 62,261 | events: avoid cloning listeners array on every emit | This PR does 2 improvements:
- Do not clone the listeners on each emit, clone when we're either adding or removing listeners
- Even when we're adding or removing listeners, if we're not currently emitting, avoid cloning again
Results:
Branch:
```sh
./node benchmark/run.js --filter ee-emit events
events/e... | null | 7ffbb768fe4f0b90d61f3293a2ef09ba7337c578 | null | low | [
{
"filename": "lib/events.js",
"patch": "@@ -87,6 +87,7 @@ const { addAbortListener } = require('internal/events/abort_listener');\n const kCapture = Symbol('kCapture');\n const kErrorMonitor = Symbol('events.errorMonitor');\n const kShapeMode = Symbol('shapeMode');\n+const kEmitting = Symbol('events.emitti... |
electron/electron | 50,590 | build(deps): bump github/codeql-action from 4.34.1 to 4.35.1 | Bumps [github/codeql-action](https://github.com/github/codeql-action) from 4.34.1 to 4.35.1.
<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>v4.35.1</h2>
<ul>
<li>Fix incorrect minimu... | null | e5b20a11d24ac39a12c75d15e204f307abcec419 | null | low | [
{
"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@38697555549f1db7851b81482ff19f1fa5c4fedc # v3.29.5\n+ ... |
ollama/ollama | 14,362 | cmd/config: install web search plugin to user-level extensions dir | The previous approach installed the plugin into the openclaw npm package's extensions directory. On WSL, files on /mnt/c appear as mode 777 (world-writable), causing openclaw's security check to block all plugins from that path.
| 34452233116171391f770e8079002620d0a9fd7a | 06edabdde1174bd31c1d93f3d9e5efc2d9504806 | 2 | medium | [
{
"filename": "cmd/config/openclaw.go",
"patch": "@@ -608,16 +608,16 @@ func clearSessionModelOverride(primary string) {\n \n const webSearchNpmPackage = \"@ollama/openclaw-web-search\"\n \n-// ensureWebSearchPlugin installs the openclaw-web-search extension into the OpenClaw\n-// extensions directory if it... |
huggingface/transformers | 44,931 | fix(camembert): add tie_word_embeddings=True to CamembertConfig | ## What does this PR do?
Fixes a v5 regression where `CamembertForMaskedLM` (and all CamemBERT masked-LM tasks) produces near-zero, near-uniform logits, making the model completely non-functional.
### Root cause
In v5, `modeling_utils.get_expanded_tied_weights_keys()` gates all weight tying on `config.tie_word_embed... | null | 9dc8d8aa3090ab3f39e6086d02b712f9274bc795 | null | low | [
{
"filename": "src/transformers/models/camembert/configuration_camembert.py",
"patch": "@@ -60,6 +60,7 @@ class CamembertConfig(PreTrainedConfig):\n classifier_dropout: float | int | None = None\n is_decoder: bool = False\n add_cross_attention: bool = False\n+ tie_word_embeddings: bool = True... |
vercel/next.js | 91,529 | Unflake two `"use cache"` tests | [Flakiness metric](https://app.datadoghq.com/ci/test/runs?query=test_level%3Atest%20%40git.repository.id%3A%22github.com%2Fvercel%2Fnext.js%22%20%40test.name%3A%28%22use-cache%20should%20cache%20results%20for%20cached%20functions%20imported%20from%20client%20components%22%20OR%20%22use-cache%20should%20cache%20results%... | null | 89c6e73c81dc959a5a6e3d5ba74d7ced8531d9bb | null | low | [
{
"filename": "test/e2e/app-dir/use-cache/use-cache.test.ts",
"patch": "@@ -188,7 +188,9 @@ describe('use-cache', () => {\n })\n \n await browser.elementById('reset-button').click()\n- expect(await browser.elementByCss('p').text()).toBe('0 0 0')\n+ await retry(async () => {\n+ expect(awai... |
huggingface/transformers | 44,971 | [ `vllm x v5`] nit | # What does this PR do?
Removed the tokenizer_class attr was never there to begin with, and kwargs are now supported.
This was failing some test on vllm ci. Fixes https://buildkite.com/vllm/ci/builds/57601/steps/canvas?sid=019d1aec-aa5a-41db-bac6-4f42397279d5
```python
In [6]: pip = pipeline("text-generation", "X... | d4895f0810fd57bf5ee8cf65c3fe20d2f622cd0a | 692d187dc7b1091717bfe941785d272f7d575563 | 7 | medium | [
{
"filename": "src/transformers/configuration_utils.py",
"patch": "@@ -32,7 +32,6 @@\n from .generation.configuration_utils import GenerationConfig\n from .modeling_gguf_pytorch_utils import load_gguf_checkpoint\n from .modeling_rope_utils import RotaryEmbeddingConfigMixin\n-from .tokenization_utils_base im... |
nodejs/node | 62,453 | crypto: runtime-deprecate DEP0203 and DEP0204 | Refs: https://github.com/nodejs/node/issues/55293
Refs: https://github.com/nodejs/node/issues/62321 | d4fa60cf9fcdcd528aac441994e0863e0eacd45b | adac0774843839cce25d526942b448c28aa53777 | 21 | medium | [
{
"filename": "doc/api/crypto.md",
"patch": "@@ -2255,12 +2255,20 @@ be listed in the `transferList` argument.\n \n <!-- YAML\n added: v15.0.0\n+changes:\n+ - version: REPLACEME\n+ pr-url: https://github.com/nodejs/node/pull/62453\n+ description: Passing a non-extractable CryptoKey as `key` is deprec... |
facebook/react | 35,672 | [DevTools] Fix commit index reset when switching profiler roots | Fixes https://github.com/facebook/react/issues/31463, https://github.com/facebook/react/issues/30114.
When switching between roots in the profiler flamegraph, the commit index was preserved from the previous root. This caused an error "Invalid commit X. There are only Y commits." when the new root had fewer commits ... | null | 5dad2b47b87d93db51f3bd41453a39124127b75d | null | low | [
{
"filename": "packages/react-devtools-shared/src/__tests__/profilerContext-test.js",
"patch": "@@ -776,6 +776,84 @@ describe('ProfilerContext', () => {\n document.body.removeChild(profilerContainer);\n });\n \n+ it('should reset commit index when switching to a different root', async () => {\n+ c... |
rust-lang/rust | 154,313 | Init `self_decl` with a correct visibility | Fixes rust-lang/rust#154295
r? petrochenkov | null | 85ddb8e4142d0cd8e34281c0c1ae2a2d496c6e55 | null | low | [
{
"filename": "compiler/rustc_resolve/src/lib.rs",
"patch": "@@ -547,6 +547,13 @@ impl ModuleKind {\n ModuleKind::Def(.., name) => name,\n }\n }\n+\n+ fn opt_def_id(&self) -> Option<DefId> {\n+ match self {\n+ ModuleKind::Def(_, def_id, _) => Some(*def_id),\n+ ... |
electron/electron | 50,594 | fix: add missing HandleScope in contentTracing.getTraceBufferUsage() | Backport of #50556
See that PR for details.
Notes: Fixed a crash when calling `contentTracing.getTraceBufferUsage()` while a trace session is active.
| null | 90decd4eaf308fd219033c88c9af5208ff820565 | null | low | [
{
"filename": "shell/browser/api/electron_api_content_tracing.cc",
"patch": "@@ -151,7 +151,10 @@ void OnTraceBufferUsageAvailable(\n gin_helper::Promise<gin_helper::Dictionary> promise,\n float percent_full,\n size_t approximate_count) {\n- auto dict = gin_helper::Dictionary::CreateEmpty(promi... |
ollama/ollama | 14,346 | mlx: remove noisy error output from dynamic library loading | The recent change in #14322 added tryLoadByName() which attempts to load libmlxc.dylib via rpath before searching directories. This is an optimization for Homebrew installations where rpath is correctly set.
However, when rpath isn't set (which is the common case for app bundle installations), dlopen fails and the C... | null | 8b4e5a82a8dc3aadfa385eb0ff0641a8c9cbbfad | null | low | [
{
"filename": "x/mlxrunner/mlx/dynamic.c",
"patch": "@@ -18,7 +18,9 @@\n \n static int mlx_dynamic_open(mlx_dynamic_handle* handle, const char* path) {\n handle->ctx = (void*) DLOPEN(path);\n- CHECK(handle->ctx != NULL);\n+ if (handle->ctx == NULL) {\n+ return 1;\n+ }\n return 0;\n }... |
electron/electron | 50,235 | build: add patch conflict resolution workflow with CI artifacts | #### Description of Change
This PR adds documentation and CI automation to streamline the patch conflict resolution workflow when fixing conflicts on branches with open PRs.
**Changes:**
1. **CLAUDE.md**: Added a new section documenting how to use CI-generated patch artifacts to resolve conflicts faster than run... | 1f6dbb09172fce997f1d088583ad320ecc92002a | 816e5964fb574585840ec82f7b1e3e99b3f93785 | 4 | medium | [
{
"filename": ".github/workflows/apply-patches.yml",
"patch": "@@ -71,3 +71,10 @@ jobs:\n uses: ./src/electron/.github/actions/checkout\n with:\n target-platform: linux\n+ - name: Upload Patch Conflict Fix\n+ if: ${{ failure() }}\n+ uses: actions/upload-artifact@bbbca2ddaa5d... |
vercel/next.js | 91,506 | fix allowedDevOrigins for no-cors requests | This PR makes configured `allowedDevOrigins` apply to cross-site no-cors dev asset requests. When browsers omit Origin for subresource loads, the dev guard now falls back to `Referer` so explicit allowlisted hosts can load `/_next/*` resources in development.
Previously, when `allowedDevOrigins` was configured, cros... | 862f9b9bb41d235e0d8cf44aa811e7fd118cee2a | d0a0474d3dc152ddeadcb2ffb05185e2a07815c9 | 28 | medium | [
{
"filename": "packages/next/src/server/lib/router-utils/block-cross-site-dev.ts",
"patch": "@@ -31,6 +31,19 @@ function warnOrBlockRequest(\n return true\n }\n \n+function parseHostnameFromHeader(\n+ header: string | string[] | undefined\n+): string | undefined {\n+ const headerValue = Array.isArray(he... |
huggingface/transformers | 44,886 | LwDetrImageLoss: Fix dtype casting to prevent crash when using amp on cuda device | # What does this PR do?
Prevent crash in class LwDetrImageLoss when using it with float16 automatic mixed precision on a Cuda device.
torch.pow causes an autocast to float32 when used with Cuda, which caused a type mismatch at ``` pos_weights[pos_ind] = t ```.
This is fixed by casting back to the original type aft... | null | 18f88de7f575410f3ad8a64418abf2572fecb259 | null | low | [
{
"filename": "src/transformers/loss/loss_lw_detr.py",
"patch": "@@ -112,6 +112,7 @@ def loss_labels(self, outputs, targets, indices, num_boxes):\n if \"logits\" not in outputs:\n raise KeyError(\"No logits were found in the outputs\")\n source_logits = outputs[\"logits\"]\n+ ... |
nodejs/node | 62,343 | src: handle null backing store in ArrayBufferViewContents::Read | added fallback stack_storage for null scenario
Fixes: https://github.com/nodejs/node/issues/62342 | null | bdf75a6c4e107595868b798b696a96ffe2c8c0e6 | null | low | [
{
"filename": "src/util-inl.h",
"patch": "@@ -591,7 +591,9 @@ void ArrayBufferViewContents<T, S>::Read(v8::Local<v8::ArrayBufferView> abv) {\n static_assert(sizeof(T) == 1, \"Only supports one-byte data at the moment\");\n length_ = abv->ByteLength();\n if (length_ > sizeof(stack_storage_) || abv->Has... |
rust-lang/rust | 154,094 | add neon load/store assembly test | <!-- homu-ignore:start -->
*[View all comments](https://triagebot.infra.rust-lang.org/gh-comments/rust-lang/rust/pull/154094)*
<!-- homu-ignore:end -->
I'm adding this test because it was requested for the beta backport of https://github.com/rust-lang/rust/issues/153336. We'd like to test this with Miri, but currently... | null | 898b62f590e802ad2fc3a6d86bb8342ab0cf8fdf | null | low | [
{
"filename": "src/tools/compiletest/src/directives/directive_names.rs",
"patch": "@@ -213,6 +213,7 @@ pub(crate) const KNOWN_DIRECTIVE_NAMES: &[&str] = &[\n \"only-apple\",\n \"only-arm\",\n \"only-arm64ec\",\n+ \"only-armv7-unknown-linux-gnueabihf\",\n \"only-avr\",\n \"only-beta\",... |
huggingface/transformers | 44,972 | [AMD CI] Gemma3/Gemma3n Expectations | As per title. Updating Gemma3/Gemma3n expectations. | a8732d5546d84bfb4519b6dbf461c947a5de45f6 | b94fee049373eb2e140d4866847559c4c42b3d7e | 4 | medium | [
{
"filename": "tests/models/gemma3/test_modeling_gemma3.py",
"patch": "@@ -453,7 +453,7 @@ def test_model_4b_bf16(self):\n (\"xpu\", 3): ['user\\nYou are a helpful assistant.\\n\\n\\n\\n\\n\\nWhat is shown in this image?\\nmodel\\nCertainly! \\n\\nThe image shows a brown cow standing on a sa... |
vercel/next.js | 91,507 | block disallowed dev origins by default | This removes the warn-only default behavior and enforces the dev-origin guard by default. Cross-origin requests to internal dev resources now block unless they match the built-in local allowlist or an explicit `allowedDevOrigins` entry. The tests are expanded to cover default blocking, configured-but-not-allowlisted ho... | d0a0474d3dc152ddeadcb2ffb05185e2a07815c9 | b2b802c043f83b047276df48d93b76d3c742f240 | 4 | medium | [
{
"filename": "docs/01-app/03-api-reference/05-config/01-next-config-js/allowedDevOrigins.mdx",
"patch": "@@ -5,11 +5,11 @@ description: Use `allowedDevOrigins` to configure additional origins that can re\n \n {/* The content of this doc is shared between the app and pages router. You can use the `<PagesOnl... |
ollama/ollama | 14,322 | mlx: try loading library via rpath before searching directories | ## Problem
The existing code in `x/mlxrunner/mlx/dynamic.go` manually searches directories for `libmlxc.*` and passes full paths to `dlopen`, bypassing the binary's rpath. This means MLX libraries installed via package managers (e.g., Homebrew) aren't found even when rpath is correctly set at link time.
## Solution
... | d18dcd77755b55c9d761e483abee17d1e2b6c58c | 458dd1b9d98a5981d1ed4134258e07186b219a69 | 16 | medium | [
{
"filename": "x/mlxrunner/mlx/dynamic.go",
"patch": "@@ -55,6 +55,30 @@ func tryLoadFromDir(dir string) bool {\n \treturn false\n }\n \n+// tryLoadByName attempts to load the library using just its name,\n+// allowing the system to use rpath, LD_LIBRARY_PATH, or standard search paths.\n+// Returns true if ... |
facebook/react | 35,590 | [react-dom] Include `submitter` in `submit` events |
## Summary
[`SubmitEvent.submitter` is widely supported](https://developer.mozilla.org/en-US/docs/Web/API/SubmitEvent/submitter#browser_compatibility) so we should expose it. We assume `submitter` is supported since https://github.com/facebook/react/pull/29028.
This doesn't include a polyfill for older browse... | null | 7b023d7073aa700b48674ae1985235b8c4755880 | null | low | [
{
"filename": "packages/react-dom-bindings/src/events/SyntheticEvent.js",
"patch": "@@ -522,6 +522,17 @@ export const SyntheticPointerEvent: $FlowFixMe = createSyntheticEvent(\n PointerEventInterface,\n );\n \n+/**\n+ * @interface SubmitEvent\n+ * @see https://html.spec.whatwg.org/multipage/form-control-i... |
huggingface/transformers | 44,674 | Officially launch parse_response | We've had `parse_response()` in the library for a while, but it's been a soft launch / prototype feature. This PR cleans it up and documents it, making it an official feature!
The API is largely unchanged from the prototype, but we drop `x-mapping` and `x-mapping-regex` since it's unclear if we need them, and they a... | a9532bcf887b364bfedcc33a13be2b1d1acbef85 | 15b3b8924b393afa37bdc36a3206eb5032bc6a66 | 12 | medium | [
{
"filename": "docs/source/en/_toctree.yml",
"patch": "@@ -124,6 +124,8 @@\n title: Tool use\n - local: chat_templating_writing\n title: Writing a chat template\n+ - local: chat_response_parsing\n+ title: Response parsing\n title: Chat with models\n - isExpanded: false\n se... |
electron/electron | 50,578 | build: upload patch conflict fix as CI artifact | Backport of #50235 and #50251
See those PRs for details. CLAUDE.md hunk dropped — that file doesn't exist on this branch.
Notes: none | null | c8be8adebf329428663f68a3269a3b878c4dc919 | null | low | [
{
"filename": ".github/workflows/apply-patches.yml",
"patch": "@@ -71,3 +71,11 @@ jobs:\n uses: ./src/electron/.github/actions/checkout\n with:\n target-platform: linux\n+ - name: Upload Patch Conflict Fix\n+ if: ${{ failure() }}\n+ uses: actions/upload-artifact@bbbca2ddaa5d... |
facebook/react | 35,176 | [react-dom] Fire onReset when automatically resetting forms | Fixes https://github.com/facebook/react/issues/33630
During the mutation phase, we disabled the event emitter. This caused `onReset` to be dropped when we [automatically reset the form when a Form Action finishes](https://github.com/facebook/react/pull/28804). Controlled inputs would not have been reset if they're r... | 875b06489f4436125613535dfe0833cd12a500f9 | dcab44d757dee6fa8d40db56715732e891a3b729 | 7 | medium | [
{
"filename": "fixtures/dom/src/components/Header.js",
"patch": "@@ -88,6 +88,7 @@ class Header extends React.Component {\n <option value=\"/mouse-events\">Mouse Events</option>\n <option value=\"/selection-events\">Selection Events</option>\n <option value=\"... |
vercel/next.js | 91,507 | block disallowed dev origins by default | This removes the warn-only default behavior and enforces the dev-origin guard by default. Cross-origin requests to internal dev resources now block unless they match the built-in local allowlist or an explicit `allowedDevOrigins` entry. The tests are expanded to cover default blocking, configured-but-not-allowlisted ho... | d0a0474d3dc152ddeadcb2ffb05185e2a07815c9 | b2b802c043f83b047276df48d93b76d3c742f240 | 4 | medium | [
{
"filename": "docs/01-app/03-api-reference/05-config/01-next-config-js/allowedDevOrigins.mdx",
"patch": "@@ -5,11 +5,11 @@ description: Use `allowedDevOrigins` to configure additional origins that can re\n \n {/* The content of this doc is shared between the app and pages router. You can use the `<PagesOnl... |
vercel/next.js | 91,530 | Turbopack: Merge release-with-assertions-no-lto profile into release-with-assertions | - More profiles means more bloat to the `target` directory and more cold builds.
- There's no practical use case where we'd be okay with the performance hit of `debug-assertions` and not okay with the performance hit of `lto = false`. Neither profile was suitable for benchmarking.
- `release-with-assertions-no-lto` is ... | 668981482a3f9de97c4c1371e940e0f84972c080 | 39e705ca9fcbf417ed4381737737d40c79fc13ef | 2 | medium | [
{
"filename": "Cargo.toml",
"patch": "@@ -271,16 +271,17 @@ opt-level = \"s\"\n [profile.release.package.miette]\n opt-level = \"s\"\n \n-\n-# Use a custom profile for CI where many tests are performance sensitive but we still want the additional validation of debug-assertions\n+# Use a custom profile for C... |
nodejs/node | 62,449 | deps: V8: cherry-pick cf1bce40a5ef | Original commit message:
[wasm] Fix S128Const on big endian
Since http://crrev.com/c/2944437 globals are no longer little endian
enforced.
S128Const handling in the initializer needs to take this into account
and byte reverse values which are hard coded in little endian order.
This... | null | cc967413c9b32e4d34d74a3b758ca71bee7a7746 | null | low | [
{
"filename": "deps/v8/src/wasm/constant-expression-interface.cc",
"patch": "@@ -40,7 +40,17 @@ void ConstantExpressionInterface::S128Const(FullDecoder* decoder,\n const Simd128Immediate& imm,\n Value* result) {\n if (... |
facebook/react | 35,659 | [Perf Tracks] Handle function names that aren't strings | Closes https://github.com/facebook/react/issues/35652
Tests are not useful here since Jest seems to get in the way. You can validate manually in the Browser console though with `class Foo { static name = {} }` and
```
{
const x = [];
const cursor = () => {
return new Proxy(() => x, { get() { return cu... | 87ae75b33f71eb673c55ddb3a147a9e58b7d05bd | da64117876002bd60baa9e93cad77421bbf0293f | 6 | medium | [
{
"filename": "packages/shared/ReactPerformanceTrackProperties.js",
"patch": "@@ -253,10 +253,15 @@ export function addValueToProperties(\n return;\n }\n case 'function':\n- if (value.name === '') {\n+ const functionName = value.name;\n+ if (\n+ functionName === '' ||... |
ollama/ollama | 14,311 | install: prevent partial download script execution | Wrap script in main function so that a truncated partial download doesn't end up executing half a script. | 935a48ed1ac35a4677736ee806139ef9618501db | 9d02d1d767b075221031be19ad9695f1259b3519 | 20 | medium | [
{
"filename": "scripts/install.sh",
"patch": "@@ -2,6 +2,10 @@\n # This script installs Ollama on Linux and macOS.\n # It detects the current operating system architecture and installs the appropriate version of Ollama.\n \n+# Wrap script in main function so that a truncated partial download doesn't end\n+#... |
rust-lang/rust | 153,463 | [beta] update stage0 and cargo | Update stage0 to 1.94.0
<https://forge.rust-lang.org/release/process.html#beta-stage0-update-friday>
---
Update cargo submodule rust-lang/rust#153515
2 commits in f298b8c82da0cba538516b45b04a480fc501d4c0..e8eb8435d5cad936237a1ee798c2f983624d0825
2026-02-24 21:59:20 +0000 to 2026-03-06 12:51:00 -0600
- [be... | null | 64e2968f2cbc9662180e8ef7ffe89df870463754 | null | low | [
{
"filename": "src/tools/cargo",
"patch": "@@ -1 +1 @@\n-Subproject commit f298b8c82da0cba538516b45b04a480fc501d4c0\n+Subproject commit e8eb8435d5cad936237a1ee798c2f983624d0825",
"additions": 1,
"deletions": 1
}
] |
facebook/react | 35,654 | [tests] Fix ReactDOMAttribute-test | In https://github.com/facebook/react/pull/35646 I thought there was a bug in trusted types, but the bug is in jsdom.
For trusted types we still want to check the coersion and throw for a good dev warning, but prod will also throw becuase the browser will implicitly coerce to a string. This ensures there's no behavi... | d4d099f05bead14cc78787f97f005b00feae56f9 | 230772f99dac80be6dda9c59441fb4928612f18e | 3 | medium | [
{
"filename": "packages/react-dom/src/__tests__/ReactDOMAttribute-test.js",
"patch": "@@ -9,6 +9,13 @@\n \n 'use strict';\n \n+// Fix JSDOM. setAttribute is supposed to throw on things that can't be implicitly toStringed.\n+const setAttribute = Element.prototype.setAttribute;\n+Element.prototype.setAttribut... |
huggingface/transformers | 44,583 | fix load_best_model_checkpoint_at_end do not load the best model chec… | …kpoint when `save_strategy` is `best`
# What does this PR do?
fix load_best_model_checkpoint_at_end do not load the best model checkpoint at the end when `save_strategy` is `"best"`
Fixes # (issue)
fix load_best_model_checkpoint_at_end do not load the best model checkpoint at the end when `save_strategy` is ... | null | e22f3088dd4d53c0a218151e124d45aaee64a63c | null | low | [
{
"filename": "src/transformers/trainer.py",
"patch": "@@ -3045,7 +3045,10 @@ def _save_checkpoint(self, model: nn.Module, trial: \"optuna.Trial | dict[str, An\n output_dir = os.path.join(run_dir, checkpoint_folder)\n self.save_model(output_dir, _internal_call=True)\n \n- if self.args... |
electron/electron | 50,577 | build: upload patch conflict fix as CI artifact | Backport of #50235 and #50251
See those PRs for details. CLAUDE.md hunk dropped — that file doesn't exist on this branch.
Notes: none | null | a7e6649ffbc848791f672f890d67774b55548b12 | null | low | [
{
"filename": ".github/workflows/apply-patches.yml",
"patch": "@@ -71,3 +71,11 @@ jobs:\n uses: ./src/electron/.github/actions/checkout\n with:\n target-platform: linux\n+ - name: Upload Patch Conflict Fix\n+ if: ${{ failure() }}\n+ uses: actions/upload-artifact@bbbca2ddaa5d... |
facebook/react | 35,650 | [Flight] Restore original function name in dev, server callstacks served with `unsafe-eval` | ## Summary
Follow-up to https://github.com/facebook/react/pull/29671
React uses `eval` with React Server Components to create fake stacks (e.g. to reconstruct Server stacks). In CSP environments, `eval` may not be available unless `unsafe-eval` is set.
We already degraded gracefully without `eval` without rest... | e66ef6480ecd19c6885f2c06dec34fec1fdc0a98 | c0c37063e2cd8976d839998573012d48d303ada0 | 3 | medium | [
{
"filename": "packages/react-client/src/ReactFlightClient.js",
"patch": "@@ -3730,6 +3730,14 @@ function createFakeFunction<T>(\n fn = function (_) {\n return _();\n };\n+ // Using the usual {[name]: _() => _()}.bind() trick to avoid minifiers\n+ // doesn't work here since this will pro... |
vercel/next.js | 92,177 | disable bmi2 in qfilter | Fixes #91708
set the legacy_x86_64_support feature on qfilter to support older intel cpus | null | 5ef10c67f52ed96d5b4e26e4dc0ea2225c96c738 | null | low | [
{
"filename": "turbopack/crates/turbo-persistence/Cargo.toml",
"patch": "@@ -26,7 +26,8 @@ memmap2 = \"0.9.5\"\n nohash-hasher = { workspace = true }\n parking_lot = { workspace = true }\n pot = \"3.0.0\"\n-qfilter = { version = \"0.3.0-alpha.2\", features = [\"serde\"] }\n+# See https://github.com/vercel/n... |
facebook/react | 35,649 | [Perf Tracks] Use minus (`-`) instead of en dash for removed props | Stacked on https://github.com/facebook/react/pull/35648
The [en dash is used for number ranges or connections](https://www.merriam-webster.com/grammar/em-dash-en-dash-how-to-use).
This seems not suitable for a diff. Our hydration diffs also use a minus symbol instead of en dash for removed lines. git-diff also us... | ff191f24b5f49252672ac4d3c46ef1e79cf7290d | 87ae75b33f71eb673c55ddb3a147a9e58b7d05bd | 1 | medium | [
{
"filename": "packages/react-reconciler/src/__tests__/ReactPerformanceTrack-test.js",
"patch": "@@ -231,7 +231,7 @@ describe('ReactPerformanceTracks', () => {\n properties: [\n ['Changed Props', ''],\n [' data', ''],\n- ['– buffer', 'null'],\n... |
ollama/ollama | 14,122 | cmd: set codex env vars on launch and handle zstd request bodies | The Codex runner was not setting OPENAI_BASE_URL or OPENAI_API_KEY, this prevents Codex from sending requests to api.openai.com instead of the local Ollama server. This mirrors the approach used by the Claude runner.
Codex v0.98.0 sends zstd-compressed request bodies to the /v1/responses endpoint in some cases. Add ... | null | 1a636fb47af789da25dd13f1cac2d4eb3b674c75 | null | low | [
{
"filename": "cmd/config/codex.go",
"patch": "@@ -6,6 +6,7 @@ import (\n \t\"os/exec\"\n \t\"strings\"\n \n+\t\"github.com/ollama/ollama/envconfig\"\n \t\"golang.org/x/mod/semver\"\n )\n \n@@ -32,6 +33,10 @@ func (c *Codex) Run(model string, args []string) error {\n \tcmd.Stdin = os.Stdin\n \tcmd.Stdout = ... |
electron/electron | 50,543 | ci: update nick-fields/retry to v4.0.0 | Backport of #50521
See that PR for details.
Notes: <!-- 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 -->none
| null | 5ae78bd5edc55d1186c42e5108a6e02625f52c28 | null | low | [
{
"filename": ".github/actions/restore-cache-azcopy/action.yml",
"patch": "@@ -24,7 +24,7 @@ runs:\n # The cache will always exist here as a result of the checkout job\n # Either it was uploaded to Azure in the checkout job for this commit\n # or it was uploaded in the checkout job for a previou... |
nodejs/node | 62,475 | lib: fix typo in esm worker loader hook comment | <!--
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#commit-message-guidelines
For code changes:
1. Inc... | null | 7ed2fdf3faefee63b0ccd5e65a0734e6096093cf | null | low | [
{
"filename": "lib/internal/modules/esm/worker.js",
"patch": "@@ -37,7 +37,7 @@ const { isCascadedLoaderInitialized, getOrInitializeCascadedLoader } = require('\n const { AsyncLoaderHooksOnLoaderHookWorker } = require('internal/modules/esm/hooks');\n \n /**\n- * Register asynchronus module loader customizat... |
huggingface/transformers | 44,930 | fix: split MXFP4 dependency checks for specific error messages | ## Summary
- Fixes #44912 — MXFP4 quantization error messages combine `is_triton_available()` and `is_kernels_available()` into a single `kernels_available` boolean, making it impossible to identify which dependency is missing
- Split the combined check into separate `triton_available` and `kernels_installed` variable... | null | 82106705eb34848411be8dd12e61c98b9224e979 | null | low | [
{
"filename": "src/transformers/quantizers/quantizer_mxfp4.py",
"patch": "@@ -90,44 +90,61 @@ def validate_environment(self, *args, **kwargs):\n \n if torch.xpu.is_available():\n is_device_supported_mxfp4 = True\n- kernels_available = is_triton_available(\"3.5.0\") and is_kern... |
ollama/ollama | 14,302 | cmd/tui: default to single-select for editor integrations | Change the multi-select model picker to start in single-select mode. Users pick one model which gets added to their existing config, with the latest pick becoming the default. Press Tab to toggle into full multi-select editing mode.
- Add singleMode/singleAdd fields to multiSelectorModel
- Tab toggles between singl... | 2f4de1acf769d4f3e1156e8a8e1059051d7f2fad | 325b72bc311b6b2b834c1a423347038ebeca31df | 5 | medium | [
{
"filename": "cmd/tui/selector.go",
"patch": "@@ -415,6 +415,12 @@ type multiSelectorModel struct {\n \tcancelled bool\n \tconfirmed bool\n \twidth int\n+\n+\t// multi enables full multi-select editing mode. The zero value (false)\n+\t// shows a single-select picker where Enter adds the chosen... |
rust-lang/rust | 147,400 | TryFrom<integer> for bool | I noticed the manual TryFrom implementation in [this video from RustConf 2025](https://youtu.be/cOP219vkax8?list=PL2b0df3jKKiRFEuVNk76ufXagOgEJ9sBZ&t=1917) and thought I'd add it. | null | e6e2d391367cb19ceb7eceb4559470a01beb177a | null | low | [
{
"filename": "library/core/src/convert/num.rs",
"patch": "@@ -39,63 +39,52 @@ impl_float_to_int!(f32 => u8, u16, u32, u64, u128, usize, i8, i16, i32, i64, i12\n impl_float_to_int!(f64 => u8, u16, u32, u64, u128, usize, i8, i16, i32, i64, i128, isize);\n impl_float_to_int!(f128 => u8, u16, u32, u64, u128, u... |
ollama/ollama | 14,290 | bugfix: better mlx model scheduling | This change properly loads in different mlx (safetensors) based models so that subsequent model loads don't improperly do inference on the already loaded model. | d18dcd77755b55c9d761e483abee17d1e2b6c58c | d07e4a1dd39c6334184c05e1c3b8192865e114c2 | 9 | medium | [
{
"filename": "server/routes_generate_test.go",
"patch": "@@ -2371,6 +2371,26 @@ func TestImageGenerateStreamFalse(t *testing.T) {\n \t\treturn nil\n \t}\n \n+\t// Create model manifest with image capability\n+\tn := model.ParseName(\"test-image\")\n+\tcfg := model.ConfigV2{Capabilities: []string{\"image\"}... |
vercel/next.js | 92,142 | Update Rspack development test manifest | This auto-generated PR updates the development integration test manifest used when testing Rspack. | null | 71be148d5f7201167033dd07d93ae510f46a1d87 | null | low | [
{
"filename": "test/rspack-dev-tests-manifest.json",
"patch": "@@ -1193,11 +1193,14 @@\n },\n \"test/development/app-dir/server-hmr/server-hmr.test.ts\": {\n \"passed\": [\n+ \"server-hmr CLI/config conflict warning should warn when CLI flag conflicts with config\",\n+ \"server-hmr metadat... |
ollama/ollama | 14,299 | cmd: ollama launch always show model picker | Based on feedback, let's let the user pick the model each time on launch | ec95c45f701523cb98795c655f8cbef212e184e6 | 2f4de1acf769d4f3e1156e8a8e1059051d7f2fad | 1 | medium | [
{
"filename": "cmd/cmd.go",
"patch": "@@ -57,9 +57,9 @@ import (\n \n func init() {\n \t// Override default selectors to use Bubbletea TUI instead of raw terminal I/O.\n-\tconfig.DefaultSingleSelector = func(title string, items []config.ModelItem) (string, error) {\n+\tconfig.DefaultSingleSelector = func(ti... |
facebook/react | 28,000 | Convert ReactFreshMultipleRenderer to createRoot | null | 4c58fc2ad8f1055f9b543147b283e5a3f795fd27 | null | low | [
{
"filename": "packages/react-refresh/src/__tests__/ReactFreshMultipleRenderer-test.internal.js",
"patch": "@@ -16,7 +16,8 @@ if (__DEV__) {\n ReactFreshRuntime = require('react-refresh/runtime');\n ReactFreshRuntime.injectIntoGlobalHook(global);\n }\n-const ReactDOM = require('react-dom');\n+const Reac... | |
electron/electron | 50,574 | perf: enable V8 builtins PGO | Backport of #50416
See that PR for details.
Notes: Enabled profile-guided optimization for V8 builtins in release builds, improving JavaScript builtin performance (Array, String, RegExp, etc.). | null | ba551d265c10545bd0b137f76fe260370008ba06 | null | low | [
{
"filename": ".github/actions/build-electron/action.yml",
"patch": "@@ -125,6 +125,9 @@ runs:\n fi\n sed $SEDOPTION '/.*builtins-pgo/d' out/Default/mksnapshot_args\n sed $SEDOPTION '/--turbo-profiling-input/d' out/Default/mksnapshot_args\n+ sed $SEDOPTION '/--reorder-builtins... |
huggingface/transformers | 44,934 | Fix failing `T5ModelIntegrationTest` | # What does this PR do?
Fixes this failing [T5ModelIntegrationTest](https://github.com/huggingface/transformers/actions/runs/23230643883/job/67524758706#step:14:1449) & this [Qwen2IntegrationTest](https://github.com/huggingface/transformers/actions/runs/23230643883/job/67524756545#step:14:1084)
<img width="2259" he... | null | a8732d5546d84bfb4519b6dbf461c947a5de45f6 | null | low | [
{
"filename": "tests/models/qwen2/test_modeling_qwen2.py",
"patch": "@@ -183,11 +183,13 @@ def test_model_450m_long_prompt_sdpa(self):\n \n @slow\n def test_speculative_generation(self):\n- EXPECTED_TEXT_COMPLETION = (\n- \"My favourite condiment is 100% natural, organic, gluten-fr... |
nodejs/node | 62,335 | quic: remove CryptoKey support from session keys option | Let's not spread any more uses of `CryptoKey` instances that would ignore its intended use properties such as algorithm and usages.
Refs: https://github.com/nodejs/node/issues/55293
Refs: https://github.com/nodejs/node/pull/62321 | null | cf2b05cab867568f344ed58829f656162685454c | null | low | [
{
"filename": "doc/api/quic.md",
"patch": "@@ -1197,9 +1197,13 @@ True to enable TLS keylogging output.\n \n <!-- YAML\n added: v23.8.0\n+changes:\n+ - version: REPLACEME\n+ pr-url: https://github.com/nodejs/node/pull/62335\n+ description: CryptoKey is no longer accepted.\n -->\n \n-* Type: {KeyObjec... |
ollama/ollama | 14,294 | cmd/config: ollama launch cline CLI | Add cline CLI, and show both under more | 948de6bbd2d07e1050d91c2422c3ee7af2d591da | ec95c45f701523cb98795c655f8cbef212e184e6 | 9 | medium | [
{
"filename": "cmd/config/cline.go",
"patch": "@@ -0,0 +1,123 @@\n+package config\n+\n+import (\n+\t\"context\"\n+\t\"encoding/json\"\n+\t\"errors\"\n+\t\"fmt\"\n+\t\"os\"\n+\t\"os/exec\"\n+\t\"path/filepath\"\n+\n+\t\"github.com/ollama/ollama/envconfig\"\n+)\n+\n+// Cline implements Runner and Editor for t... |
facebook/react | 35,634 | [Flight] Fix markup types for server references | Follow-up to https://github.com/facebook/react/pull/35632/changes/d5324f41438fd58d01d5a5c2d6e4775872775b41
Flow narrows the `ServerReferenceId` to a `string` due to `typeof` check and then it can't assign the narrowed value to an opaque type.
Instead we don't make `ServerReferenceId` opaque like for the other con... | 10680271fab565e0edf948d3a6dc9d30e83df94c | 8c34556ca8df0dab34bbaf68e0dd15cd4a5e3f7f | 1 | medium | [
{
"filename": "packages/react-client/src/forks/ReactFlightClientConfig.markup.js",
"patch": "@@ -18,7 +18,7 @@ export * from 'react-client/src/ReactClientConsoleConfigPlain';\n export type ModuleLoading = null;\n export type ServerConsumerModuleMap = null;\n export opaque type ServerManifest = null;\n-expor... |
rust-lang/rust | 154,614 | Streamline `CachingSourceMapView` | `CachingSourceMapView` is home to some pretty ugly, low-level, repetitive code. We can do better. Details in individual commits.
r? @JonathanBrouwer | null | 8931f23941cf9ec0093f86a81337873b288344c9 | null | low | [
{
"filename": "compiler/rustc_middle/src/query/on_disk_cache.rs",
"patch": "@@ -231,7 +231,7 @@ impl OnDiskCache {\n type_shorthands: Default::default(),\n predicate_shorthands: Default::default(),\n interpret_allocs: Default::default(),\n- sour... |
vercel/next.js | 92,081 | Improve error message for deprecated `experimental.dynamicIO` config | ### What?
Added a specific, actionable error message when users have `experimental.dynamicIO` in their `next.config` file. Instead of the generic "invalid experimental key" message, users now see that `dynamicIO` has been replaced by `cacheComponents`, with a link to the documentation.
### Why?
When `experimen... | 9a0214a12f1ed1901c4b01e84073ab854848eea0 | 9b66a376e7e66c51d1a0952a96e3c377b56bd81a | 19 | medium | [
{
"filename": "packages/next/src/server/config.ts",
"patch": "@@ -84,6 +84,12 @@ function normalizeNextConfigZodErrors(\n \"\\nUse 'experimental.turbopackFileSystemCacheForDev' instead.\"\n message +=\n '\\nLearn more: https://nextjs.org/docs/app/api-reference/config/next-config-... |
nodejs/node | 62,438 | tools: bump picomatch in /tools/doc | Bumps and [picomatch](https://github.com/micromatch/picomatch). These dependencies needed to be updated together.
Updates `picomatch` from 2.3.1 to 2.3.2
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a href="https://github.com/micromatch/picomatch/releases">picomatch's releases</a>.</em></p>
<blockqu... | 583c479c9f2ffe1d8ca489282023838ab19e632b | 9289ad5cbbb46414cc0ee98d7a9026edc5160e15 | 12 | medium | [
{
"filename": "tools/doc/package-lock.json",
"patch": "@@ -2328,29 +2328,6 @@\n \"hast-util-to-html\": \"^9.0.5\"\n }\n },\n- \"node_modules/@shikijs/engine-javascript\": {\n- \"version\": \"3.23.0\",\n- \"resolved\": \"https://registry.npmjs.org/@shikijs/engine-javascript/-/e... |
facebook/react | 35,629 | [flags] make enableComponentPerformanceTrack static everywhere | Follow-up to https://github.com/facebook/react/pull/34665.
Already gated on `enableProfilerTimer` everywhere, which is only enabled for `__PROFILE__`, except for Flight should be unified in a future. | 94913cbffeff60c862ef818ca62cdc329dc2fd57 | 699abc89cea33d5aa0b443050ae27c2fea7cb3cb | 12 | medium | [
{
"filename": "packages/shared/forks/ReactFeatureFlags.native-fb-dynamic.js",
"patch": "@@ -25,4 +25,3 @@ export const passChildrenWhenCloningPersistedNodes = __VARIANT__;\n export const enableFragmentRefs = __VARIANT__;\n export const enableFragmentRefsScrollIntoView = __VARIANT__;\n export const enableFra... |
electron/electron | 50,573 | perf: enable V8 builtins PGO | Backport of #50416
See that PR for details.
Notes: Enabled profile-guided optimization for V8 builtins in release builds, improving JavaScript builtin performance (Array, String, RegExp, etc.). | e7d473337fec3af8843c44bf5616783b77f668fe | d2e0d1998517c88f18c74279dec4ce915a76f00d | 21 | medium | [
{
"filename": ".github/actions/build-electron/action.yml",
"patch": "@@ -128,6 +128,9 @@ runs:\n fi\n sed $SEDOPTION '/.*builtins-pgo/d' out/Default/mksnapshot_args\n sed $SEDOPTION '/--turbo-profiling-input/d' out/Default/mksnapshot_args\n+ sed $SEDOPTION '/--reorder-builtins... |
vercel/next.js | 92,167 | Add rust-fingerprint task and SCCACHE passthrough env | ## What
Replace duplicated Rust input globs across every turbo task with a single rust-fingerprint task that computes a TURBO_HASH stamp file. All Rust build/check tasks now depend on this stamp instead of repeating the full glob list.
Also add `SCCACHE_*` to `globalPassThroughEnv` so `sccache` env vars are avail... | 8a241bb504bf167d191b96e914b7785156510038 | 4b62ff78c0ad4117477ab78806df809f4c1f8b33 | 1 | medium | [
{
"filename": "packages/next-swc/package.json",
"patch": "@@ -16,6 +16,7 @@\n \"build-native-wasi\": \"npx --package=@napi-rs/cli@3.0.0-alpha.45 napi build --platform --target wasm32-wasip1-threads -p next-napi-bindings --cwd ../../ --output-dir packages/next-swc/native --no-default-features\",\n \"... |
vercel/next.js | 91,968 | next.config.js: Accept an option for serverFastRefresh | ### What?
Adds `experimental.serverFastRefresh` to `next.config.js` so users (and Next.js plugins) can opt out of server-side Fast Refresh without needing to pass a CLI flag.
Also adds a `--server-fast-refresh` positive CLI flag (hidden) alongside the existing `--no-server-fast-refresh`, and emits a warning when ... | c43965275556c3e0243de822e5bd21e07cf9d48c | a21f3f4008a7d4c1778649d533de76b62dffabaa | 1 | medium | [
{
"filename": "packages/next/src/bin/next.ts",
"patch": "@@ -314,7 +314,14 @@ program\n '--experimental-https-ca, <path>',\n 'Path to a HTTPS certificate authority file.'\n )\n- .option('--no-server-fast-refresh', 'Disable server-side Fast Refresh')\n+ // `--server-fast-refresh` is hidden becaus... |
huggingface/transformers | 44,895 | Add static FP8 expert support | # What does this PR do?
This PR enables static FP8 experts. This also works on multi-gpu with device-map. A fix for that was to set was to set `torch.cuda.set_device()`. Triton's JIT compiler uses he active device context to determine which GPU to compile and launch the kernel on. If tensors are on GPU:1 but the a... | null | e5ad3946209fb96db5e9965b3eb67d69cc3749e0 | null | low | [
{
"filename": "src/transformers/integrations/finegrained_fp8.py",
"patch": "@@ -186,7 +186,29 @@ def w8a8_fp8_matmul(\n if _supports_cutlass(A, B, block_size, output_dtype):\n return w8a8_block_fp8_matmul_cutlass(A, B, As, Bs, output_dtype)\n \n- # Fall back to Triton\n+ # Ensure correct C... |
ollama/ollama | 14,289 | bugfix: add missing linear layer factory | null | 3a88f7eb2087170a0835dd2a1bdb2ddad4e5cdd5 | null | low | [
{
"filename": "x/mlxrunner/model/linear.go",
"patch": "@@ -0,0 +1,92 @@\n+//go:build mlx\n+\n+package model\n+\n+import (\n+\t\"github.com/ollama/ollama/x/mlxrunner/mlx\"\n+\t\"github.com/ollama/ollama/x/models/nn\"\n+)\n+\n+// LinearFactory builds linear layers using shared tensor maps and quant defaults.\... | |
facebook/react | 35,599 | [DevTools] Update inspected element on component filter changes | ## Summary
We used to only update the inspected element when selected host element changes. This is fine when Components and the browser's Elements panel are alternate. With the inspected element pane you can now have the inspected element and browsers Elements panel side-by-side.
Now we update the inspected elem... | null | 2c30ebc4e397d57fe75f850e32aa44e353052944 | null | low | [
{
"filename": "packages/react-devtools-shared/src/backend/agent.js",
"patch": "@@ -1086,6 +1086,15 @@ export default class Agent extends EventEmitter<{\n }\n renderer.updateComponentFilters(componentFilters);\n }\n+\n+ // Due to the component filters changing, we might be able\n+ ... |
nodejs/node | 62,371 | repl: use vm DONT_CONTEXTIFY context | REPL does not need interception behavior in the vm context. Creating
a `DONT_CONTEXTIFY` context to reduce the chance to be impacted by V8
interceptor API changes.
No observable behavior or tests should break. This does not fix the
behavior in `vm` with the V8 upgrade, only REPL is fixed.
Refs: https://github.... | null | 1ea93c7f3ac5c495cb8c113c0f4c995589acc9c5 | null | low | [
{
"filename": "lib/repl.js",
"patch": "@@ -1132,10 +1132,10 @@ class REPLServer extends Interface {\n session.once('Runtime.executionContextCreated', ({ params }) => {\n this[kContextId] = params.context.id;\n });\n- context = vm.createContext();\n+ context = vm.creat... |
electron/electron | 50,408 | feat: add accessibilityDisplayShouldDifferentiateWithoutColor on macOS | Backport of #49912
See that PR for details.
Notes: Added nativeTheme.shouldDifferentiateWithoutColor on macOS | null | 6a8d187105d830ca863eef7b2bd2313078c140b4 | null | low | [
{
"filename": "docs/api/native-theme.md",
"patch": "@@ -84,3 +84,7 @@ Currently, Windows high contrast is the only system setting that triggers forced\n ### `nativeTheme.prefersReducedTransparency` _Readonly_\n \n A `boolean` that indicates whether the user has chosen via system accessibility settings to re... |
rust-lang/rust | 150,752 | Update libc to v0.2.183 | <!-- homu-ignore:start -->
*[View all comments](https://triagebot.infra.rust-lang.org/gh-comments/rust-lang/rust/pull/150752)*
<!-- homu-ignore:end -->
<!-- homu-ignore:start -->
<!--
If this PR is related to an unstable feature or an otherwise tracked effort,
please link to the relevant tracking issue here. If you... | null | 470facce550e4b479e30e112f9994b40954d115f | null | low | [
{
"filename": "library/Cargo.lock",
"patch": "@@ -146,9 +146,9 @@ dependencies = [\n \n [[package]]\n name = \"libc\"\n-version = \"0.2.178\"\n+version = \"0.2.183\"\n source = \"registry+https://github.com/rust-lang/crates.io-index\"\n-checksum = \"37c93d8daa9d8a012fd8ab92f088405fb202ea0b6ab73ee2482ae66af4... |
huggingface/transformers | 44,909 | Fix: Update optimization.py | The `get_scheduler` function was identifying the `inverse_sqrt` scheduler type but failing to pass `**scheduler_specific_kwargs` to the underlying `get_inverse_sqrt_schedule` function.
This caused user-defined parameters like `timescale` to be silently ignored. This commit adds the missing kwargs to the function ca... | null | 2f121eaf3c1b3692e49a10172a24714dce7858a1 | null | low | [
{
"filename": "src/transformers/optimization.py",
"patch": "@@ -1031,7 +1031,7 @@ def scheduler_hook(param):\n return schedule_func(optimizer, num_warmup_steps=num_warmup_steps)\n \n if name == SchedulerType.INVERSE_SQRT:\n- return schedule_func(optimizer, num_warmup_steps=num_warmup_step... |
vercel/next.js | 92,163 | [test] Skip flaky `segment-cache-refresh` suite | See https://app.datadoghq.com/ci/test/runs?query=test_level%3Atest%20%40test.source.file%3Atest%2Fe2e%2Fapp-dir%2Fsegment-cache%2Frefresh%2Fsegment-cache-refresh.test.ts&agg_m=count&agg_m_source=base&agg_t=count&citest_explorer_sort=timestamp%2Cdesc&cols=%40test.status%2Ctimestamp%2C%40duration%2C%40test.type%2C%40test... | null | b7f99c773ed73e8916bd14ebc9709f625220c838 | null | low | [
{
"filename": "test/e2e/app-dir/segment-cache/refresh/segment-cache-refresh.test.ts",
"patch": "@@ -2,7 +2,8 @@ import { nextTestSetup } from 'e2e-utils'\n import type * as Playwright from 'playwright'\n import { createRouterAct } from 'router-act'\n \n-describe('segment cache (refresh)', () => {\n+// Disab... |
ollama/ollama | 14,285 | bugfix: display the parameter count correctly in mlx for ollama show | a0407d07fa900a129afcec42fc222a19f1102d82 | 0d5da826d47f6ca219f04302234b77a849476fba | 27 | medium | [
{
"filename": "x/server/show.go",
"patch": "@@ -5,6 +5,7 @@ import (\n \t\"encoding/json\"\n \t\"fmt\"\n \t\"io\"\n+\t\"math\"\n \t\"os\"\n \t\"sort\"\n \t\"strings\"\n@@ -58,7 +59,15 @@ func GetSafetensorsLLMInfo(name model.Name) (map[string]any, error) {\n \t\t}\n \t}\n \n-\treturn buildModelInfo(config, ... | |
ollama/ollama | 14,277 | model: add llama3 architecture to mlxrunner | This change builds off PR #14276 and adds the llama3 architecture to the mlxrunner. | 041fb7763945d11e0484cb5c838b5fea67741c94 | 9b795698b8fe99369ca7dc208f8368b3ba47b918 | 1 | medium | [
{
"filename": "x/mlxrunner/imports.go",
"patch": "@@ -5,4 +5,5 @@ package mlxrunner\n import (\n \t_ \"github.com/ollama/ollama/x/models/gemma3\"\n \t_ \"github.com/ollama/ollama/x/models/glm4_moe_lite\"\n+\t_ \"github.com/ollama/ollama/x/models/llama\"\n )",
"additions": 1,
"deletions": 0
},
{
... |
huggingface/transformers | 44,858 | [CB] [Minor] Simplify test suite | Right now, the continuous batching tests all use similar mechanisms, namely:
1. loading a model and a tokenizer
2. preparing data for generate or generate_batch
3. running generate to compare its outputs with generate_batch
This PR adds 3 helper functions to do that, increasing code re-usability across the file.
... | 19bba39fa79e6ad126d2e95c593cb51d71ca7241 | 4f8de83424c066d2edb412876501731dbb62f88a | 5 | medium | [
{
"filename": "tests/generation/test_continuous_batching.py",
"patch": "@@ -16,6 +16,7 @@\n import gc\n import itertools\n import unittest\n+from typing import Any\n from unittest.mock import patch\n \n import torch\n@@ -28,6 +29,7 @@\n CompileConfig,\n ContinuousBatchingConfig,\n GenerationConf... |
nodejs/node | 62,325 | zlib: fix use-after-free when reset() is called during write | The `Reset()` method did not check the `write_in_progress_` flag before resetting the compression stream. This allowed `reset()` to free the compression library's internal state while a worker thread was still using it during an async write, causing a use-after-free.
Add a `write_in_progress_` guard to `Reset()` tha... | null | 53bcd114b10021c4a883b08df4d3c2ff6946b430 | null | low | [
{
"filename": "src/node_zlib.cc",
"patch": "@@ -644,6 +644,12 @@ class CompressionStream : public AsyncWrap,\n CompressionStream* wrap;\n ASSIGN_OR_RETURN_UNWRAP(&wrap, args.This());\n \n+ if (wrap->write_in_progress_) {\n+ wrap->env()->ThrowError(\n+ \"Cannot reset zlib stream whil... |
ollama/ollama | 14,271 | Update: download link in README for macOS (#1) | 519b11eba198576e1c6f2f60155420323e4edf1a | 8224cce583e6e7253e2fdeee8f07ab4c8da7bce5 | 8 | medium | [
{
"filename": "README.md",
"patch": "@@ -16,7 +16,7 @@ Start building with open models.\n curl -fsSL https://ollama.com/install.sh | sh\n ```\n \n-or [download manually](http://localhost:8080/download/Ollama.dmg)\n+or [download manually](https://ollama.com/download/Ollama.dmg)\n \n ### Windows\n ",
"add... | |
rust-lang/rust | 153,018 | `unused_must_use` lint improvements | <!-- homu-ignore:start -->
<!--
If this PR is related to an unstable feature or an otherwise tracked effort,
please link to the relevant tracking issue here. If you don't know of a related
tracking issue or there are none, feel free to ignore this.
This PR will get automatically assigned to a reviewer. In case y... | null | 0fc8bb5beb6d5508e8515f2d76bbda925971f49f | null | low | [
{
"filename": "compiler/rustc_lint/src/lib.rs",
"patch": "@@ -78,7 +78,7 @@ mod transmute;\n mod types;\n mod unit_bindings;\n mod unqualified_local_imports;\n-mod unused;\n+pub mod unused;\n mod utils;\n \n use async_closures::AsyncClosureUsage;\n@@ -125,6 +125,7 @@ use transmute::CheckTransmutes;\n use ty... |
facebook/react | 35,494 | [Fiber] Correctly handle replaying when hydrating | When hydrating if something suspends and then resolves in a microtask it is possible that React will resume the render without fully unwinding work in progress. This can cause hydration cursors to be offset and lead to hydration errors. This change adds a restore step when replaying HostComponent to ensure the hydratio... | null | c18662405cc436646411647f8a8965c1c0594c3c | null | low | [
{
"filename": "packages/react-dom/src/__tests__/ReactDOMServerPartialHydration-test.internal.js",
"patch": "@@ -4063,4 +4063,170 @@ describe('ReactDOMServerPartialHydration', () => {\n expect(span.style.display).toBe('');\n expect(ref.current).toBe(span);\n });\n+\n+ // Regression for https://git... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.