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 |
|---|---|---|---|---|---|---|---|---|
ggml-org/llama.cpp | 20,529 | ggml : add native AVX512-FP16 support for F16 operations | The overall benchmark speed remains almost the same because the CPU is now calculating faster than the RAM can deliver the data. (See perf stat results below showing 2.7 billion fewer instructions).
Also note that this path will be only enabled for native build or with custom flags.
now:
```
Performance counte... | null | d0b79aaa2f6e7b7d3c26b1845b43cef158697540 | null | low | [
{
"filename": "ggml/src/ggml-cpu/simd-mappings.h",
"patch": "@@ -479,13 +479,51 @@ do { \\\n \n // F16 AVX512\n \n-// F16 AVX\n+#if defined(__AVX512FP16__)\n+\n+#define GGML_F16_STEP 128\n+#define GGML_F16_EPR 32\n+\n+#define GGML_F16x32 ... |
huggingface/transformers | 43,356 | Update ASR, audio classification, and TTS documentation | Update ASR, audio classification, and TTS documentation to include additional library dependencies and correct dataset references. Added 'report_to' parameter for tracking in training examples. | null | 9055ee4dd9ae6e258b8244faccfbdfa5c8e313e4 | null | low | [
{
"filename": "docs/source/en/tasks/asr.md",
"patch": "@@ -36,7 +36,7 @@ To see all architectures and checkpoints compatible with this task, we recommend\n Before you begin, make sure you have all the necessary libraries installed:\n \n ```bash\n-pip install transformers datasets evaluate jiwer\n+pip instal... |
ollama/ollama | 9,788 | fix: correctly save in interactive mode | This fixes the case where a FROM line in previous modelfile points to a file which may/may not be present in a different ollama instance. We shouldn't be relying on the filename though and instead just check if the FROM line was instead a valid model name and point to that instead.
Fixes #9573 | 4bed739259d8d4c4c0f79f2359f13fbbd3249ece | 2c8b4846437747bd23e7a176f83011e39ec2128b | 18 | medium | [
{
"filename": "cmd/cmd_test.go",
"patch": "@@ -757,3 +757,132 @@ func TestCreateHandler(t *testing.T) {\n \t\t})\n \t}\n }\n+\n+func TestNewCreateRequest(t *testing.T) {\n+\ttests := []struct {\n+\t\tname string\n+\t\tfrom string\n+\t\topts runOptions\n+\t\texpected *api.CreateRequest\n+\t}{\n+\... |
ggml-org/llama.cpp | 15,526 | vulkan: enable Conv2D for Apple after MoltenVK fixed the bug | The bug that caused Conv2D shader compile failure on MoltenVK has been fixed in 1.4.0. Remove the workaround for it. | 611f419cff11e4952228162a1c44cb35dff2274a | a9c6ffcbfacee092bfaaa400306fceda18199737 | 3 | high | [
{
"filename": "ggml/src/ggml-vulkan/ggml-vulkan.cpp",
"patch": "@@ -11853,14 +11853,13 @@ static bool ggml_backend_vk_device_supports_op(ggml_backend_dev_t dev, const ggm\n // Op is disabled for Apple because it segfaults at pipeline create time on MoltenVK\n ggml_backend_vk_... |
electron/electron | 49,551 | test: remove split dependency | #### 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
-->
Removes the `split` dependency as we can just use `node:readline` to get the same end ... | null | a7de47084bbd03e921df58a2e50ac01e72fb597a | null | low | [
{
"filename": "spec/api-app-spec.ts",
"patch": "@@ -2,7 +2,6 @@ import { app, BrowserWindow, Menu, session, net as electronNet, WebContents, uti\n \n import { assert, expect } from 'chai';\n import * as semver from 'semver';\n-import split = require('split')\n \n import * as cp from 'node:child_process';\n ... |
ollama/ollama | 9,775 | server/internal/client/ollama: set User-Agent for registry client | This sets the agent header in DefaultRegistry to include the version of
the client, OS, and architecture in the previous format, with a minor
twist.
Note: The version is obtained from the build info, instead of the
version in version.Version, which should not longer be necessary, but we
can remove in a future co... | 4e320b8b90b8a698fc3c057a3f54cbabe59b543a | 8294676150dde3dd3d316644b86e42b07204ff9c | 7 | medium | [
{
"filename": "server/internal/client/ollama/registry.go",
"patch": "@@ -25,6 +25,7 @@ import (\n \t\"os\"\n \t\"path/filepath\"\n \t\"runtime\"\n+\t\"runtime/debug\"\n \t\"slices\"\n \t\"strconv\"\n \t\"strings\"\n@@ -259,6 +260,7 @@ func DefaultRegistry() (*Registry, error) {\n \t}\n \n \tvar rc Registry\... |
nodejs/node | 61,328 | doc: add esm and cjs examples to `node:v8` | This PR adds the missing `ESM` and `CJS` examples to their respective counterparts for the [V8 documentation](https://nodejs.org/api/v8.html).
For the [`setFlagsFromString`](https://nodejs.org/api/v8.html#v8setflagsfromstringflags) example, we weren't printing anything back to the user so the example looked like we ... | null | a49053ad6df32c414058242409d442872161dfec | null | low | [
{
"filename": "doc/api/v8.md",
"patch": "@@ -7,7 +7,11 @@\n The `node:v8` module exposes APIs that are specific to the version of [V8][]\n built into the Node.js binary. It can be accessed using:\n \n-```js\n+```mjs\n+import v8 from 'node:v8';\n+```\n+\n+```cjs\n const v8 = require('node:v8');\n ```\n \n@@ ... |
facebook/react | 33,579 | Make it clearer what runtime release failed | I thought somebody was experimenting with another release. This hopefully reduces future confusion.
```diff
-Publish of $stable release failed
+[Runtime] Publish of stable@next,canary release failed | d60f77a533da830613431ddef83d0eda928697ad | c8822e926b79205fd2e828c81e031bd0afc0effc | 29 | medium | [
{
"filename": ".github/workflows/runtime_prereleases.yml",
"patch": "@@ -108,5 +108,5 @@ jobs:\n with:\n webhook-url: ${{ secrets.DISCORD_WEBHOOK_URL }}\n embed-author-name: \"GitHub Actions\"\n- embed-title: 'Publish of $${{ inputs.release_channel }} release failed'\n+ ... |
vercel/next.js | 88,821 | [test] Skip failing deploy test in `searchparams-reuse-loading.test.ts` | `searchparams-reuse-loading With Middleware should correctly return different RSC data for full prefetches with different searchParam values` has been failing for the past 3 months which was hidden by not running the test consistently across retries.
Skipping it unless we find out why until Thursday.
Part of http... | null | 2cfc7ebb0065f3a016cf2a73cb24d9865f7c656f | null | low | [
{
"filename": "test/deploy-tests-manifest.json",
"patch": "@@ -41,6 +41,11 @@\n \"app dir - metadata react cache should have same title and page value when navigating\"\n ]\n },\n+ \"test/e2e/app-dir/searchparams-reuse-loading/searchparams-reuse-loading.test.ts\": {\n+ \"failed\": ... |
vercel/next.js | 88,826 | [test] Skip flaky `prefetch-runtime` tests for deploy tests | These are [all the prefetch-runtime tests that caused e2e release deploy test to fail ultimately](https://app.datadoghq.com/ci/test/runs?query=test_level%3Atest%20%40ci.pipeline.name%3Atest-e2e-deploy-release%20%40ci.pipeline.url%3A%2A%2Fattempts%2F2%20%40test.status%3Afail%20%40test.source.file%3Atest%2Fe2e%2Fapp-dir%... | 2cfc7ebb0065f3a016cf2a73cb24d9865f7c656f | 5a0b9787c7ef8beb5bc194c5e7182112919f0567 | 1 | medium | [
{
"filename": "test/deploy-tests-manifest.json",
"patch": "@@ -11,6 +11,43 @@\n \"app dir client cache semantics (experimental staleTimes) static: 180 prefetch={undefined} - default should re-use the loading boundary for the custom static override time (3 minutes)\"\n ]\n },\n+ \"test/e... |
huggingface/transformers | 38,856 | [Bugfix][informer]: Correct tensor shape for input_size=1 | Hi @Rocketknight1, thanks for the great guidance on the previous PR!
This new pull request follows your suggestion to fix the bug at its source. It resolves a RuntimeError that occurs in time series models inheriting from TimeSeriesTransformerModel (such as InformerModel) when config.input_size is set to 1.
The r... | null | 334bf913dca3f8f85312d264c0f7da5607384ced | null | low | [
{
"filename": "src/transformers/models/informer/modeling_informer.py",
"patch": "@@ -1461,8 +1461,14 @@ def create_network_inputs(\n )\n \n # static features\n- log_abs_loc = loc.abs().log1p() if self.config.input_size == 1 else loc.squeeze(1).abs().log1p()\n- log_scale = scale... |
ggml-org/llama.cpp | 14,770 | Vulkan: Fix fprintf format-security warning | Fixes #14745 | null | 83f5872404baa39d826af2ef66351e63c64205a8 | null | low | [
{
"filename": "ggml/src/ggml-vulkan/vulkan-shaders/vulkan-shaders-gen.cpp",
"patch": "@@ -765,8 +765,8 @@ void write_output_files() {\n len += \"};\\n\";\n }\n }\n- fprintf(src, data.c_str());\n- fprintf(src, len.c_str());\n+ fputs(data.c_str(), src);... |
vuejs/vue | 9,346 | update README.md | <!--
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 | ec8403209ab9ecc526f696c032f6cd14c8d122bb | null | low | [
{
"filename": "dist/README.md",
"patch": "@@ -40,7 +40,7 @@ module.exports = {\n }\n }\n }\n-````\n+```\n \n #### Rollup\n ",
"additions": 1,
"deletions": 1
}
] |
electron/electron | 49,557 | test: remove split dependency | Backport of #49551
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 -->
| null | cd00dcbafdceafc18a9b2c4881f9baea488a7fda | null | low | [
{
"filename": "spec/api-app-spec.ts",
"patch": "@@ -2,7 +2,6 @@ import { app, BrowserWindow, Menu, session, net as electronNet, WebContents, uti\n \n import { assert, expect } from 'chai';\n import * as semver from 'semver';\n-import split = require('split')\n \n import * as cp from 'node:child_process';\n ... |
ollama/ollama | 9,776 | fix: gemma3 quantization | This change allows users to use `ollama create --quantize <level>` to quantize gemma3 models.
We still use llama.cpp in order to do quantization with `ollama create` and do not make calls directly to ggml (as with `ollama run`) so we need to define the architecture so that llama.cpp can correctly do the quantization... | null | ef378ad673a3f01382add316835957b1d4184177 | null | low | [
{
"filename": "llama/llama.cpp/src/llama-arch.cpp",
"patch": "@@ -37,6 +37,7 @@ static const std::map<llm_arch, const char *> LLM_ARCH_NAMES = {\n { LLM_ARCH_MINICPM3, \"minicpm3\" },\n { LLM_ARCH_GEMMA, \"gemma\" },\n { LLM_ARCH_GEMMA2, \"gemma2\"... |
vercel/next.js | 88,558 | Turbopack: MappedReadRef should be Send and Sync | ### What?
To be able to used MappedReadRefs in async context, they need to be Send. | 125b7d89f497e04770cdb4abb11d7e3734b60afc | 43460b17ec2594b26493bcf5906857412cb833fd | 13 | medium | [
{
"filename": "turbopack/crates/turbo-tasks/src/lib.rs",
"patch": "@@ -56,11 +56,11 @@ mod invalidation;\n mod join_iter_ext;\n mod key_value_pair;\n pub mod keyed;\n-pub mod keyed_read_ref;\n #[doc(hidden)]\n pub mod macro_helpers;\n mod magic_any;\n mod manager;\n+pub mod mapped_read_ref;\n mod marker_tra... |
facebook/react | 33,560 | build: make enableComponentPerformanceTrack dynamic for native-fb | ## Summary
Make this flag dynamic, so it can be controlled internally.
## How did you test this change?
Build, observe that `console.timeStamp` is only present in FB artifacts and `enableComponentPerformanceTrack` is referenced.
| 5d24c64cc9c019fc644c4c6f0da640131b80ba18 | 374dfe8edf8beef62e5bb312f99c600a232f353f | 25 | medium | [
{
"filename": "packages/shared/forks/ReactFeatureFlags.native-fb-dynamic.js",
"patch": "@@ -26,3 +26,4 @@ export const passChildrenWhenCloningPersistedNodes = __VARIANT__;\n export const enableLazyPublicInstanceInFabric = __VARIANT__;\n export const renameElementSymbol = __VARIANT__;\n export const enableFr... |
facebook/react | 33,571 | [compiler] Cleanup debugging code |
Removes unnecessary debugging code in the new inference passes now that they've stabilized more.
---
[//]: # (BEGIN SAPLING FOOTER)
Stack created with [Sapling](https://sapling-scm.com). Best reviewed with [ReviewStack](https://reviewstack.dev/facebook/react/pull/33571).
* __->__ #33571
* #33558
* #33547 | d37faa041bce86c1cbb05fdbc839440c9d9de9cf | 2bee34867d30083ce01232baccb72b6fa696456b | 1 | medium | [
{
"filename": "compiler/packages/babel-plugin-react-compiler/src/Inference/InferMutationAliasingEffects.ts",
"patch": "@@ -57,7 +57,6 @@ import {\n import {\n printAliasingEffect,\n printAliasingSignature,\n- printFunction,\n printIdentifier,\n printInstruction,\n printInstructionValue,\n@@ -194,... |
huggingface/transformers | 43,348 | fix: Correct ESM TrunkConfig self-modulo in validation | ### What does this PR do?
`TrunkConfig` validation contained a self-modulo check that would never trigger (dead code). This PR corrects the divisibility check to use the head width instead.
Fixes #43347.
### Before submitting
* [x] This PR fixes a typo or improves the docs (you can dismiss the other checks ... | aec03e5d1921038bdcb0bd8915e696ca02c82983 | e9d68f6a03914847e3ee73389f0415f684fc0e88 | 13 | medium | [
{
"filename": "src/transformers/models/esm/configuration_esm.py",
"patch": "@@ -101,16 +101,6 @@ def __post_init__(self):\n \n if self.max_recycles <= 0:\n raise ValueError(f\"`max_recycles` should be positive, got {self.max_recycles}.\")\n- if self.sequence_state_dim % self.seque... |
nodejs/node | 61,339 | deps: update sqlite to 3.51.2 | This is an automated update of sqlite to 3.51.2. | null | cab20f2396ed2c38b0001642c49d998354d70eb2 | null | low | [
{
"filename": "deps/sqlite/sqlite3.c",
"patch": "@@ -1,6 +1,6 @@\n /******************************************************************************\n ** This file is an amalgamation of many separate C source files from SQLite\n-** version 3.51.1. By combining all the individual C code files into this\n+** v... |
vercel/next.js | 88,559 | Turbopack: selective read for module_chunk_groups in ChunkGroupInfo | ### What?
Use selective reads for module chunk groups from chunk group info
This is a preparation to make chunking be less dependent on module graph changes. | cfe29f98d8d7cdb9295a588653dfa430e7247013 | bbf8aec99b4fb127f8265b402df24a82acccab9e | 7 | medium | [
{
"filename": "turbopack/crates/turbopack-core/src/chunk/chunking/mod.rs",
"patch": "@@ -280,6 +280,7 @@ async fn batch_chunk_items_with_info_with_type(\n }\n \n /// Creates chunks based on heuristics for the passed `chunk_items`.\n+#[tracing::instrument(level = Level::TRACE, skip_all)]\n pub async fn make_... |
facebook/react | 33,558 | [compiler] Preserve Create effects, guarantee effects initialize once |
Ensures that effects are well-formed with respect to the rules:
* For a given instruction, each place is only initialized once (w one of Create, CreateFrom, Assign)
* Ensures that Alias targets are already initialized within the same instruction (should have a Create before them)
* Preserves Create and similar instruc... | 7ceb10035faf79eb6a645938ab15099126714381 | d37faa041bce86c1cbb05fdbc839440c9d9de9cf | 4 | medium | [
{
"filename": "compiler/packages/babel-plugin-react-compiler/src/Inference/AnalyseFunctions.ts",
"patch": "@@ -20,10 +20,11 @@ import {inferReactiveScopeVariables} from '../ReactiveScopes';\n import {rewriteInstructionKindsBasedOnReassignment} from '../SSA';\n import {inferMutableRanges} from './InferMutabl... |
vercel/next.js | 88,560 | Turbopack: use selective reads for merged modules | ### What?
Use selective reads for merged modules data.
This is a preparation to make chunking be less dependent on module graph changes. | bbf8aec99b4fb127f8265b402df24a82acccab9e | a8b0f9db70b6e89d3d31090853bcf5a0c8dc59a9 | 1 | medium | [
{
"filename": "turbopack/crates/turbopack-core/src/chunk/chunk_group.rs",
"patch": "@@ -1,4 +1,4 @@\n-use std::{cell::RefCell, collections::HashSet, sync::atomic::AtomicBool};\n+use std::{collections::HashSet, sync::atomic::AtomicBool};\n \n use anyhow::{Context, Result};\n use rustc_hash::FxHashMap;\n@@ -3... |
vercel/next.js | 88,561 | Turbopack: add caching to chunking | ### What?
Make make_chunks a cached function to allow it to be cached even when the module graph changes (assuming that chunk group's chunk items are not affected by the change)
| a8b0f9db70b6e89d3d31090853bcf5a0c8dc59a9 | c38e57fd239f140020d4b44e0b1afb29e021db66 | 1 | medium | [
{
"filename": "turbopack/crates/turbopack-browser/src/chunking_context.rs",
"patch": "@@ -695,6 +695,8 @@ impl ChunkingContext for BrowserChunkingContext {\n )\n .await?;\n \n+ let chunks = chunks.await?;\n+\n let mut assets = chunks\n .iter()\n... |
ggml-org/llama.cpp | 14,249 | Vulkan: Fix host-pinned memory for large allocations | This must have slipped through when the max size was added. Next step to "modernize" the host buffer system would be to make it device-specific, although I'm not sure if that would do anything currently. | c89c2d1ab94b11845240b7d3313c87691ea18d88 | 10bb545c5b54175ed9874ad8d187effa2bcb4b5f | 27 | medium | [
{
"filename": "ggml/src/ggml-vulkan/ggml-vulkan.cpp",
"patch": "@@ -9495,6 +9495,12 @@ static size_t ggml_backend_vk_host_buffer_type_get_alignment(ggml_backend_buffer\n UNUSED(buft);\n }\n \n+static size_t ggml_backend_vk_host_buffer_type_get_max_size(ggml_backend_buffer_type_t buft) {\n+ return vk_... |
ollama/ollama | 9,635 | Align versions for local builds | Darwin was using a different pattern for the version string than linux or windows. | null | 2d2247e59e3995c00bf6bdf9bd2713bdf01f6921 | null | low | [
{
"filename": "scripts/build_darwin.sh",
"patch": "@@ -8,7 +8,7 @@ usage() {\n exit 1\n }\n \n-export VERSION=${VERSION:-$(git describe --tags --dirty)}\n+export VERSION=${VERSION:-$(git describe --tags --first-parent --abbrev=7 --long --dirty --always | sed -e \"s/^v//g\")}\n export GOFLAGS=\"'-ldflags... |
electron/electron | 49,316 | test: fix flaky `BrowserView` test | #### Description of Change
Chromium applies a `#121212` background color in Dark mode, while in Light mode the background is transparent or white when loading `about:blank`.
As a result, the test fails when the desktop theme is set to Dark.
<!--
Thank you for your Pull Request. Please provide a description ... | null | 82d350524e181241b2437e914a794c5ab13a48f0 | null | low | [
{
"filename": "spec/api-browser-view-spec.ts",
"patch": "@@ -90,7 +90,7 @@ describe('BrowserView module', () => {\n w.show();\n w.setBounds(display.bounds);\n w.setBackgroundColor(WINDOW_BACKGROUND_COLOR);\n- await w.loadURL('about:blank');\n+ await w.loadURL('data:text/html,<htm... |
vuejs/vue | 9,324 | [feature] allow template to be function in vue-server-renderer | <!--
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 | b65f6d78e0e480601b0042b1b5e8259343b629fb | null | low | [
{
"filename": "src/server/create-renderer.js",
"patch": "@@ -23,7 +23,7 @@ export type RenderOptions = {\n directives?: Object;\n isUnaryTag?: Function;\n cache?: RenderCache;\n- template?: string;\n+ template?: string | (content: string, context: any) => string;\n inject?: boolean;\n basedir?: ... |
facebook/react | 33,532 | [compiler] Rename InferFunctionExprAliasingEffectsSignature | ---
[//]: # (BEGIN SAPLING FOOTER)
Stack created with [Sapling](https://sapling-scm.com). Best reviewed with [ReviewStack](https://reviewstack.dev/facebook/react/pull/33532).
* #33571
* #33558
* #33547
* #33543
* #33533
* __->__ #33532
* #33530 | 7c28c154651f54a63a6823e4dc54faa218c05a89 | 7ceb10035faf79eb6a645938ab15099126714381 | 10 | medium | [
{
"filename": "compiler/packages/babel-plugin-react-compiler/src/Inference/AnalyseFunctions.ts",
"patch": "@@ -22,7 +22,7 @@ import {inferMutableRanges} from './InferMutableRanges';\n import inferReferenceEffects from './InferReferenceEffects';\n import {assertExhaustive} from '../Utils/utils';\n import {in... |
ggml-org/llama.cpp | 12,434 | Vulkan: Default to 1GB allocations instead of 4GB to avoid fragmentation and driver issues | Extend the changes done in #11551 to all cases to avoid driver allocation issues that keep coming up. On some drivers allocations fail for other reasons than fragmentation, despite being below the size limit. Some examples in #5441. This should fix that.
I couldn't reproduce performance differences I saw in the prev... | 7dfad387e3f6ac98d383ded2d175eb59736a3993 | fd123cfead49eb32e386e26b8ef7a6d41554dda5 | 2 | high | [
{
"filename": "ggml/src/ggml-vulkan/ggml-vulkan.cpp",
"patch": "@@ -2524,13 +2524,9 @@ static vk_device ggml_vk_get_device(size_t idx) {\n \n if (GGML_VK_SUBALLOCATION_BLOCK_SIZE != nullptr) {\n device->suballocation_block_size = std::stoul(GGML_VK_SUBALLOCATION_BLOCK_SIZE);\n-#if define... |
electron/electron | 49,556 | test: remove split dependency | Backport of #49551
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 -->
| 75ee26902b658d74cc2e4419598446203272fc87 | ecbe8ee08a1561330b14e1bb3b38dc90c088817d | 6 | medium | [
{
"filename": "spec/api-app-spec.ts",
"patch": "@@ -2,7 +2,6 @@ import { app, BrowserWindow, Menu, session, net as electronNet, WebContents, uti\n \n import { assert, expect } from 'chai';\n import * as semver from 'semver';\n-import split = require('split')\n \n import * as cp from 'node:child_process';\n ... |
facebook/react | 33,530 | [compiler] More readable alias signature declarations |
Now that we have support for defining aliasing signatures in moduleTypeProvider, which uses string names for receiver/args/returns/etc, we can reuse that same form for builtin declarations. The declarations are written in the unparsed form and than parsed/validated when registered (in the addFunction/addHook call).
T... | 34179fe3449e141e980bbeaa8fc0a61b156113bb | 4335f699870920932c8f30b6ad52173c3d819d16 | 1 | medium | [
{
"filename": "compiler/packages/babel-plugin-react-compiler/src/HIR/Globals.ts",
"patch": "@@ -5,14 +5,7 @@\n * LICENSE file in the root directory of this source tree.\n */\n \n-import {\n- Effect,\n- GeneratedSource,\n- makeIdentifierId,\n- Place,\n- ValueKind,\n- ValueReason,\n-} from './HIR';\n+... |
huggingface/transformers | 43,346 | fix failed unit test for glm_moe_lite model | This PR fixes following failed test cases:
```
FAILED tests/models/glm4_moe_lite/test_modeling_glm4_moe_lite.py::Glm4MoeModelTest::test_eager_matches_fa2_generate
- RuntimeError: mat1 and mat2 shapes cannot be multiplied (14x256 and 512x32)
FAILED tests/models/glm4_moe_lite/test_modeling_glm4_moe_lite.py::Glm4MoeM... | null | ca99469eb1d243b5f5b3a1f7a94938d9b988c3a6 | null | low | [
{
"filename": "tests/models/glm4_moe_lite/test_modeling_glm4_moe_lite.py",
"patch": "@@ -47,13 +47,15 @@ def __init__(\n q_lora_rank=16,\n qk_nope_head_dim=64,\n qk_rope_head_dim=64,\n+ v_head_dim=128,\n ):\n super().__init__(parent=parent)\n self.n_routed_... |
nodejs/node | 60,141 | cluster: fix port reuse between cluster | Fixes: https://github.com/nodejs/node/issues/60086
If the `message.index` passed to [`queryServer`](https://github.com/nodejs/node/pull/60141/files#diff-6af1a5122ce960c2853fe60a767019d74ab5339a8e161f3ef1a13317d1fbfb57R274-R278) becomes out of sync due to errors or other issues, it may prevent a cluster from reusing ... | null | 903f64796e65324db71ddedee5216a0704f42c1a | null | low | [
{
"filename": "lib/internal/cluster/primary.js",
"patch": "@@ -271,8 +271,12 @@ function queryServer(worker, message) {\n return;\n \n const key = `${message.address}:${message.port}:${message.addressType}:` +\n- `${message.fd}:${message.index}`;\n- let handle = handles.get(key);\n+ ... |
ggml-org/llama.cpp | 20,484 | vendor : update cpp-httplib to 0.37.2 | 557fe2d9132913eaf08c8abf21b0cff61addb9ac | 77e20cc1076f351f6ba37b545a20b21c6207cbe5 | 15 | medium | [
{
"filename": "scripts/sync_vendor.py",
"patch": "@@ -5,7 +5,7 @@\n import sys\n import subprocess\n \n-HTTPLIB_VERSION = \"refs/tags/v0.37.1\"\n+HTTPLIB_VERSION = \"refs/tags/v0.37.2\"\n \n vendor = {\n \"https://github.com/nlohmann/json/releases/latest/download/json.hpp\": \"vendor/nlohmann/json.h... | |
vercel/next.js | 88,918 | remove gt workflow from agents.md | graphite workflow should be optional not required to do all changes. agents sometimes are aware of the `gt` cmd existence and force to update PR even didn't plan to push | null | d1616083289a7979738112aacef993ebe9a22368 | null | low | [
{
"filename": "AGENTS.md",
"patch": "@@ -39,62 +39,6 @@ The main Next.js framework lives in `packages/next/`. This is what gets publishe\n - `packages/font/` - `next/font` implementation\n - `packages/third-parties/` - Third-party script integrations\n \n-## Git Workflow\n-\n-**CRITICAL: Use Graphite (`gt`)... |
ollama/ollama | 9,747 | Support multiple images in Gemma3 | This fixes tensor corruption that was occurring when images were getting split across batches by giving models more control over batching. With that change, it is possible to support multiple images in a single input on Gemma3.
resolves #9697 | null | 7bf793a6007ca11fae0180ea6f2ebd7258428bd4 | null | low | [
{
"filename": "server/prompt.go",
"patch": "@@ -26,7 +26,6 @@ func chatPrompt(ctx context.Context, m *Model, tokenize tokenizeFunc, opts *api.\n \tvar system []api.Message\n \n \tisMllama := checkMllamaModelFamily(m)\n-\tisGemma3 := checkGemma3ModelFamily(m)\n \n \tvar imageNumTokens int\n \t// TODO: Ideall... |
vuejs/vue | 9,343 | [automated] Patreon sponsors update | This is an automated pull request. | null | 6ca18bd05b5f90af5285851cc6db62b908a844e2 | null | low | [
{
"filename": "BACKERS.md",
"patch": "@@ -21,7 +21,7 @@ Funds donated via Patreon go directly to support Evan You's full-time work on Vu\n <img width=\"260px\" src=\"https://raw.githubusercontent.com/vuejs/vuejs.org/master/themes/vue/source/images/stdlib.png\">\n </a>\n </p>\n-\n+ \n <!--special end-... |
vercel/next.js | 88,651 | Turbopack: Make the priority_runner testcase deterministic | # What
Fixes the `test_mixed_cpu_bound_and_waiting_tasks` to avoid flakes.
# Why
The `test_mixed_cpu_bound_and_waiting_tasks` was non-deterministic and relied on thread scheduling order.
The previous test relied on timing-based synchronization using sleep() calls with varying durations to simulate CPU-bound... | 1bdb2e20052281c614a8bfef08a6ebb66a546129 | 4674a4358739f2e203a445a33bc3672fc499f615 | 13 | medium | [
{
"filename": "turbopack/crates/turbo-tasks/src/priority_runner.rs",
"patch": "@@ -389,7 +389,11 @@ impl Future for JoinHandle {\n \n #[cfg(test)]\n mod tests {\n- use std::{sync::Arc, thread::sleep, time::Duration};\n+ use std::{\n+ sync::{Arc, Barrier},\n+ thread::sleep,\n+ time... |
electron/electron | 48,027 | feat: Add `getAccentColor` on Linux | ~~Marked as draft for now since my [upstream change](https://chromium-review.googlesource.com/c/chromium/src/+/6831464) has not yet been reviewed/approved.~~
The included patch has been [merged upstream](https://chromium-review.googlesource.com/c/chromium/src/+/6831464) in Chromium.
#### Description of Change
... | null | 7ec0ebc50a016ed5ea538160602001024d84cb06 | null | low | [
{
"filename": "docs/api/system-preferences.md",
"patch": "@@ -14,7 +14,7 @@ console.log(systemPreferences.getEffectiveAppearance())\n \n The `systemPreferences` object emits the following events:\n \n-### Event: 'accent-color-changed' _Windows_\n+### Event: 'accent-color-changed' _Windows_ _Linux_\n \n Retu... |
facebook/react | 33,522 | [compiler] Repro for case of lost precision in new inference |
In comparing compilation output of the old/new inference models I found this case (heavily distilled into a fixture). Roughly speaking the scenario is:
* Create a mutable object `x`
* Extract part of that object and pass it to a hook/jsx so that _part_ becomes frozen
* Mutate `x`, even indirectly.
In the old model w... | null | 0e7cdebb32817de0e0bbee3b362f0959e36c959c | null | low | [
{
"filename": "compiler/packages/babel-plugin-react-compiler/src/__tests__/fixtures/compiler/new-mutability/repro-jsx-captures-value-mutated-later.expect.md",
"patch": "@@ -0,0 +1,53 @@\n+\n+## Input\n+\n+```javascript\n+// @flow @enableNewMutationAliasingModel\n+\n+import {identity, Stringify, useFragment}... |
nodejs/node | 61,342 | zlib: validate write_result array length | Fixes: https://github.com/nodejs/node/issues/61286
<!--
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 | 2679b62e9155ea7d0849cd9ca5b380675aa29234 | null | low | [
{
"filename": "src/node_zlib.cc",
"patch": "@@ -736,6 +736,7 @@ class ZlibStream final : public CompressionStream<ZlibContext> {\n \n CHECK(args[4]->IsUint32Array());\n Local<Uint32Array> array = args[4].As<Uint32Array>();\n+ CHECK_GE(array->Length(), 2);\n Local<ArrayBuffer> ab = array->Buff... |
vercel/next.js | 88,668 | Turbopack: Various cleanup for turbo-tasks-fs, mostly retry logic and string formatting | - https://github.com/vercel/next.js/pull/87661 removed the `tokio::task::spawn_blocking` call from `retry_blocking`, so we can simplify things a lot by removing the `Send + 'static` bounds, and eliminate a bunch of cloning `Path`s into `PathBuf`s.
- `retry_blocking` no longer takes a path argument. After removing the p... | 917ff14ce2a1dc60913dab436e8bddb8952ae93b | f263c5622d0a48ddac1fc8e88d9f396f24a818e7 | 3 | medium | [
{
"filename": "turbopack/crates/turbo-tasks-fs/src/lib.rs",
"patch": "@@ -39,7 +39,7 @@ use std::{\n fmt::{self, Debug, Display, Formatter},\n fs::FileType,\n future::Future,\n- io::{self, BufRead, BufReader, ErrorKind, Read},\n+ io::{self, BufRead, BufReader, ErrorKind, Read, Write as _},... |
ggml-org/llama.cpp | 20,480 | opencl: fix l2_norm | Fix test-backend-ops failures with l2_norm.
| null | 3b439504ba49d41c75885d34c1353d195e09e028 | null | low | [
{
"filename": "ggml/src/ggml-opencl/kernels/l2_norm.cl",
"patch": "@@ -63,7 +63,7 @@ kernel void kernel_l2_norm_f32(\n \n barrier(CLK_LOCAL_MEM_FENCE);\n \n- const float scale = 1.0f/sqrt(max(sum[0], eps));\n+ const float scale = 1.0f/max(sqrt(sum[0]), eps);\n \n for (int i00 = get_local_id(0)... |
facebook/react | 33,518 | [compiler] Fix infinite loop due to uncached applied signatures |
When we apply new aliasing signatures we can generate new temporaries, which causes the abstract memory model to not converge. The fix is to make sure we cache the applications of these signatures.
---
[//]: # (BEGIN SAPLING FOOTER)
Stack created with [Sapling](https://sapling-scm.com). Best reviewed with [ReviewStac... | 8f4ce72f0bfe02e51e9a7c704dc33122d909f292 | 81d81151169be4b1b0ad8bd6439e94cfc982bb5a | 1 | medium | [
{
"filename": "compiler/packages/babel-plugin-react-compiler/src/Inference/AliasingEffects.ts",
"patch": "@@ -8,6 +8,7 @@\n import {CompilerErrorDetailOptions} from '../CompilerError';\n import {\n FunctionExpression,\n+ GeneratedSource,\n Hole,\n IdentifierId,\n ObjectMethod,\n@@ -18,6 +19,7 @@ im... |
huggingface/transformers | 42,681 | Fix post_process_semantic_segmentation removing valid class in Conditional DETR | Description:
Fixes #42679
What does this PR do?
Conditional DETR does not have a null class unlike DETR (its classifier outputs num_labels classes, not num_labels + 1). However, the post_process_semantic_segmentation method was incorrectly applying [..., :-1] to remove what it assumed was a null class, which actuall... | null | 27c901740e03152bd7b8fdbc52ddf32d10b86b59 | null | low | [
{
"filename": "src/transformers/models/conditional_detr/image_processing_conditional_detr.py",
"patch": "@@ -1483,7 +1483,6 @@ def post_process_object_detection(\n \n return results\n \n- # Copied from transformers.models.detr.image_processing_detr.DetrImageProcessor.post_process_semantic_segment... |
ollama/ollama | 9,741 | fix: error if image requested without vision model | this change will return an error to the user if they try to use vision on a model in the ollama engine that does not support vision or has not loaded any vision tensors
resolves #9717 | 63a394068c44149252b061a1aded05b0530868a8 | 543240fb5f0eb1a5443fd2b45f857e7dd4dcbfed | 21 | medium | [
{
"filename": "model/model.go",
"patch": "@@ -22,6 +22,8 @@ import (\n \t\"github.com/ollama/ollama/model/input\"\n )\n \n+var ErrNoVisionModel = errors.New(\"this model is missing data required for image input\")\n+\n // Model implements a specific model architecture, defining the forward pass and any mode... |
facebook/react | 33,514 | [commit] Improve error for hoisting violations |
The previous error for hoisting violations pointed only to the variable declaration, but didn't show where the value was accessed before that declaration. We now track where each hoisted variable is first accessed and report two errors, one for the reference and one for the declaration. When we improve our diagnostic ... | b067c6fe7962ce2c8ddca6deeadcb21afe2538f8 | 8f4ce72f0bfe02e51e9a7c704dc33122d909f292 | 2 | medium | [
{
"filename": "compiler/packages/babel-plugin-react-compiler/src/Inference/InferMutationAliasingEffects.ts",
"patch": "@@ -38,6 +38,7 @@ import {\n import {\n eachInstructionValueLValue,\n eachInstructionValueOperand,\n+ eachTerminalOperand,\n eachTerminalSuccessor,\n } from '../HIR/visitors';\n impo... |
huggingface/transformers | 43,327 | Getting closer | It was necessary to flatten the LoRA weights for 3d MoE, as LoRA always expected 2d weights (being nn.Linear).
Therefore, I added new conversion ops: `PermuteDims` and `FlattenDims`. I also added a convenience function `_block_diag_3d`, it could probably be optimized.
Furthermore, I now have separate branches for... | 50e4f0ef0275c9fdedba24160eb9dd5e989fb71a | 6857f5f2f3f1a0d1e8044e327c05a2380739f38a | 1 | medium | [
{
"filename": "src/transformers/integrations/peft.py",
"patch": "@@ -21,6 +21,7 @@\n from ..core_model_loading import (\n Concatenate,\n ConversionOps,\n+ MergeModulelist,\n WeightConverter,\n WeightRenaming,\n )\n@@ -56,8 +57,62 @@\n from ..modeling_utils import LoadStateDictConfig\n... |
vercel/next.js | 88,869 | Turbopack: Use webpki-root-certs in addition to rustls-platform-verifier on Linux for bare-bones Linux images without root CA stores | This should fix the user-reported issue here: https://github.com/vercel/next.js/pull/88290#issuecomment-3762940063
That issue occurs when building with bare-bones docker images that contain no root CA store.
It's inconvenient to test inside of a bare-bones docker image, but I did at least test this works on a normal ... | 8fe8ff79157a675b56259a8eaf2d16fc1647b254 | 78bc2bad8be1f7e11a8cbe76bc8175fb7ee781ae | 2 | medium | [
{
"filename": "Cargo.lock",
"patch": "@@ -610,7 +610,7 @@ dependencies = [\n \"bitflags 2.9.1\",\n \"cexpr\",\n \"clang-sys\",\n- \"itertools 0.10.5\",\n+ \"itertools 0.12.1\",\n \"lazy_static\",\n \"lazycell\",\n \"log\",\n@@ -4641,7 +4641,7 @@ version = \"0.50.3\"\n source = \"registry+https://githu... |
ggml-org/llama.cpp | 20,443 | graph : remove redundant GDN state transposes | cont #20437
fix #20436
As correctly noted in #20436, there is no need to transpose the recurrent state in the GDN computation. Simplify the ggml graph for the unfused path + optimize the fused kernels using coalesced read/writes.
TODOs:
- [x] Wait for #20334 to be merged and adapt the current PR | b5e1212063e3a605bd421c0edfee1a4a59731f65 | e30f1fdf74ea9238ff562901aa974c75aab6619b | 5 | medium | [
{
"filename": "ggml/src/ggml-cpu/ops.cpp",
"patch": "@@ -10477,34 +10477,40 @@ static void ggml_compute_forward_gated_delta_net_one_chunk(\n const float beta_val = *(const float *)((const char *)src_beta->data + iv3 * nbb3 + t * nbb2 + iv1 * nbb1);\n const float * g_d = (const fl... |
ollama/ollama | 9,703 | count gemma3 vision tensors | this also fixes a bug with the logging the memory requirements where it miscounts the layer sizes
resolves #9685 | 4bed739259d8d4c4c0f79f2359f13fbbd3249ece | 4ea4d2b18952a104a47079df00ac0bfe4706f3aa | 3 | medium | [
{
"filename": "fs/ggml/ggml.go",
"patch": "@@ -583,39 +583,52 @@ func (f GGML) GraphSize(context, batch uint64, kvCacheType string) (kv, partialO\n }\n \n func (llm GGML) VisionGraphSize() (weights, graphSize uint64) {\n-\tswitch llm.KV().Architecture() {\n-\tcase \"mllama\":\n-\t\tfor _, layer := range llm... |
vercel/next.js | 88,669 | Turbopack: Tweak retry loop for link creation to try to fix os error 80 on Windows | Though I've been unable to reproduce the issue (prior PRs in this stack are my attempts at that), I believe this *should* fix the OS Error 80 ([`ERROR_FILE_EXISTS`](https://learn.microsoft.com/en-us/windows/win32/debug/system-error-codes--0-499-)) that we've seen users report on Windows: https://github.com/vercel/next.... | f263c5622d0a48ddac1fc8e88d9f396f24a818e7 | 8fe8ff79157a675b56259a8eaf2d16fc1647b254 | 1 | medium | [
{
"filename": "Cargo.lock",
"patch": "@@ -9544,6 +9544,7 @@ dependencies = [\n \"serde_path_to_error\",\n \"sha2\",\n \"tempfile\",\n+ \"thiserror 1.0.69\",\n \"tokio\",\n \"tracing\",\n \"triomphe 0.1.12\",",
"additions": 1,
"deletions": 0
},
{
"filename": "turbopack/crates/turbo-task... |
facebook/react | 33,573 | [compiler] update fixtures | ---
[//]: # (BEGIN SAPLING FOOTER)
Stack created with [Sapling](https://sapling-scm.com). Best reviewed with [ReviewStack](https://reviewstack.dev/facebook/react/pull/33573).
* #33571
* #33558
* #33547
* #33543
* #33533
* #33532
* #33530
* #33526
* #33522
* #33518
* #33514
* __->__ #33573 | 90ccbd71c158a8aeb1bf3ec704011ddd58842b71 | 7ce2a63acc199a4f8829625470ea56a347632340 | 5 | medium | [
{
"filename": "compiler/packages/babel-plugin-react-compiler/src/__tests__/fixtures/compiler/inner-function/nullable-objects/array-map-named-callback-cross-context.expect.md",
"patch": "@@ -57,67 +57,62 @@ import { Stringify } from \"shared-runtime\";\n * - cb1 is not assumed to be called since it's only u... |
electron/electron | 49,555 | test: remove split dependency | Backport of #49551
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 -->
| 218300e57f0864940ddd091f370ab2311435f5fd | 87badb84f8a9808b42f35f0806d83ddcefd5f484 | 21 | medium | [
{
"filename": "spec/api-app-spec.ts",
"patch": "@@ -2,7 +2,6 @@ import { app, BrowserWindow, Menu, session, net as electronNet, WebContents, uti\n \n import { assert, expect } from 'chai';\n import * as semver from 'semver';\n-import split = require('split')\n \n import * as cp from 'node:child_process';\n ... |
vuejs/vue | 9,275 | fix(async component): memory leak after synchronous async loading (fix #9229) | <!--
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 | d21e93139697be2c0a6fdc4ee74d30d2834a729f | null | low | [
{
"filename": "src/core/vdom/helpers/resolve-async-component.js",
"patch": "@@ -78,6 +78,8 @@ export function resolveAsyncComponent (\n // (async resolves are shimmed as synchronous during SSR)\n if (!sync) {\n forceRender(true)\n+ } else {\n+ contexts.length = 0\n }\n ... |
huggingface/transformers | 43,305 | move tts from multimodal to audio doc | 847a68df5975649ea6ebe49969863b79c965a26b | 000438aadf3105ee8473ee29ef678e8cfbca606a | 11 | medium | [
{
"filename": "docs/source/en/_toctree.yml",
"patch": "@@ -263,6 +263,8 @@\n title: Audio classification\n - local: tasks/asr\n title: Automatic speech recognition\n+ - local: tasks/text-to-speech\n+ title: Text to speech\n title: Audio\n - sections:\n - loc... | |
facebook/react | 33,512 | [compiler] FunctionExpression context locations point to first reference |
This has always been awkward: `FunctionExpression.context` places have locations set to the declaration of the identifier, whereas other references have locations pointing to the reference itself. Here, we update context operands to have their location point to the first reference of that variable within the function.... | 66cfe048d3ab02afd3eeba9e8d7710acb3a4ab38 | e081cb344652dc3003d9194cca618292a889ff2a | 6 | medium | [
{
"filename": "compiler/packages/babel-plugin-react-compiler/src/HIR/BuildHIR.ts",
"patch": "@@ -72,21 +72,21 @@ export function lower(\n env: Environment,\n // Bindings captured from the outer function, in case lower() is called recursively (for lambdas)\n bindings: Bindings | null = null,\n- captur... |
nodejs/node | 60,854 | process: optimize asyncHandledRejections by using FixedQueue | branch:
```sh
./node benchmark/run.js --filter handled-rejections process
process/handled-rejections.js
process/handled-rejections.js n=10000: 237,537.91758393912
process/handled-rejections.js n=50000: 306,685.9060958424
process/handled-rejections.js n=100000: 324,124.4183992467
```
main:
```sh
./node... | null | 012bf70908fafcbdf8d5c819e346c32b80f7aea4 | null | low | [
{
"filename": "benchmark/process/handled-rejections.js",
"patch": "@@ -0,0 +1,43 @@\n+'use strict';\n+\n+const common = require('../common.js');\n+\n+// Benchmarks the throughput of processing many promise rejections that are\n+// initially unhandled, get warned, and then handled asynchronously, exercising\... |
huggingface/transformers | 43,321 | Replace config.get() with getattr(config) | A buggy line was introduced in #42894 that uses `config.get()`, but `config` classes are not dicts and don't have item access like that. Replaced it with `getattr()`!
Fixes #43272, cc @arthurzucker for review | aec03e5d1921038bdcb0bd8915e696ca02c82983 | 7d05b604eec12d093b85cc058fe5577d76f4cc10 | 7 | medium | [
{
"filename": "src/transformers/models/auto/tokenization_auto.py",
"patch": "@@ -716,7 +716,7 @@ def from_pretrained(\n )\n config = config.encoder\n \n- model_type = config_class_to_model_type(type(config).__name__) or config.get(\"model_type\", None)\n+ model_type... |
facebook/react | 33,504 | [commit] Better error message for invalid hoisting |
We're already tracking which variables are hoisted context variables, so if we see a mutation of a frozen value we can emit a custom error message to help users identify the problem.
---
[//]: # (BEGIN SAPLING FOOTER)
Stack created with [Sapling](https://sapling-scm.com). Best reviewed with [ReviewStack](https://revi... | 66cfe048d3ab02afd3eeba9e8d7710acb3a4ab38 | 7b67dc92b0339062ce8b6a1d64a458d7c8f04561 | 5 | medium | [
{
"filename": "compiler/packages/babel-plugin-react-compiler/src/Inference/InferMutationAliasingEffects.ts",
"patch": "@@ -901,23 +901,44 @@ function applyEffect(\n console.log(prettyFormat(state.debugAbstractValue(value)));\n }\n \n- const reason = getWriteErrorReason({\n- ... |
ggml-org/llama.cpp | 19,954 | tools : enable kvu in perplexity for hellaswag, winogrande, multiple-choice | llama-perplexity -hf unsloth/Qwen3-0.6B-GGUF:Q4_K_M -f winogrande-debiased-eval.csv --winogrande
winogrande_score : tokenizing selected tasks
winogrande_score : calculating winogrande score over selected tasks.
split_equal: sequential split is not supported when there are coupled sequences in the input... | null | 463b6a963c2de376e102d878a50d26802f15833c | null | low | [
{
"filename": "tools/perplexity/perplexity.cpp",
"patch": "@@ -2025,21 +2025,14 @@ int main(int argc, char ** argv) {\n return 1;\n }\n \n- const bool ppl = !params.hellaswag && !params.winogrande && !params.multiple_choice && !params.kl_divergence;\n-\n- if (ppl || params.kl_divergence) {... |
ollama/ollama | 9,744 | fix: change default context size for gemma3 | 63a394068c44149252b061a1aded05b0530868a8 | 80c7ce381ba5bd435503323007c4049fc1c3faef | 19 | medium | [
{
"filename": "convert/convert_gemma3.go",
"patch": "@@ -87,7 +87,7 @@ func (p *gemma3Model) KV(t *Tokenizer) ggml.KV {\n \t\tkv[\"gemma3.embedding_length\"] = p.HiddenSize\n \t\tkv[\"gemma3.feed_forward_length\"] = p.IntermediateSize\n \tdefault:\n-\t\tkv[\"gemma3.context_length\"] = cmp.Or(p.MaxPositionEm... | |
huggingface/transformers | 43,285 | Fix incorrect dimension naming in Gemma3 projector | # What does this PR do?
This PR fixes a misleading variable naming in `Gemma3MultiModalProjector.forward`.
Currently, the third dimension of the vision encoder output (which represents the **hidden size**) is assigned to a variable named `seq_length`. This is confusing because the second dimension is the actual s... | aec03e5d1921038bdcb0bd8915e696ca02c82983 | fdef96bfd4985976eaaa87e0c480ff82199ddeda | 5 | medium | [
{
"filename": "src/transformers/models/gemma3/modeling_gemma3.py",
"patch": "@@ -694,11 +694,11 @@ def __init__(self, config: Gemma3Config):\n self.avg_pool = nn.AvgPool2d(kernel_size=self.kernel_size, stride=self.kernel_size)\n \n def forward(self, vision_outputs: torch.Tensor):\n- batch... |
facebook/react | 33,500 | [compiler] Fix AnalyzeFunctions to fully reset context identifiers |
AnalyzeFunctions had logic to reset the mutable ranges of context variables after visiting inner function expressions. However, there was a bug in that logic: InferReactiveScopeVariables makes all the identifiers in a scope point to the same mutable range instance. That meant that it was possible for a later function ... | 66cfe048d3ab02afd3eeba9e8d7710acb3a4ab38 | 7c28c154651f54a63a6823e4dc54faa218c05a89 | 4 | medium | [
{
"filename": "compiler/packages/babel-plugin-react-compiler/src/Inference/AnalyseFunctions.ts",
"patch": "@@ -42,8 +42,16 @@ export default function analyseFunctions(func: HIRFunction): void {\n * Reset mutable range for outer inferReferenceEffects\n */\n for (const operand ... |
electron/electron | 49,559 | fix: chrome://accessibility drift | Backport of #49547
See that PR for details.
Notes: Fixed an issue in `chrome://accessibility`. | null | 2f3a1ca4616eb682c4320e7cb47773de9f2e1a3b | null | low | [
{
"filename": "shell/browser/ui/webui/accessibility_ui.cc",
"patch": "@@ -49,6 +49,10 @@\n #include \"ui/views/widget/widget.h\"\n #include \"ui/views/widget/widget_delegate.h\"\n \n+#if BUILDFLAG(IS_WIN)\n+#include \"ui/accessibility/platform/ax_platform_node_win.h\"\n+#endif\n+\n namespace {\n \n constexp... |
huggingface/transformers | 43,319 | Updated Ben | As discussed internally.
Still not finished, running into this error now:
```
MixtralForCausalLM LOAD REPORT from: peft-internal-testing/mixtral-pre-v5-lora
Key | Status | Details ... | 95e76243446f79c4c1478d9c472269604a17464e | 50e4f0ef0275c9fdedba24160eb9dd5e989fb71a | 1 | medium | [
{
"filename": "src/transformers/integrations/peft.py",
"patch": "@@ -11,6 +11,7 @@\n # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.\n # See the License for the specific language governing permissions and\n # limitations under the License.\n+import copy\n import inspect\n import j... |
vuejs/vue | 9,199 | fix(#9198): fix merged twice bug when passing extended constructor to mixins | (#9198)
<!-- PULL REQUEST TEMPLATE -->
<!-- (Update "[ ]" to "[x]" to check a box) -->
**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 introdu... | null | 537161779ea329c1d0a993997555f1c692b8cac1 | null | low | [
{
"filename": "src/core/util/options.js",
"patch": "@@ -382,13 +382,13 @@ export function mergeOptions (\n }\n \n if (typeof child === 'function') {\n- child = child.options\n+ child = child.extendOptions\n }\n \n normalizeProps(child, vm)\n normalizeInject(child, vm)\n normalizeDirectives... |
vercel/next.js | 88,892 | [devtools] Fix notch coloring of error overlay in forced colors mode | When chrome://flags/#enable-force-dark is enabled:
https://github.com/user-attachments/assets/a4a22f01-b7e3-4a42-b92d-1eec472dec25
When chrome://flags/#enable-force-dark is disabled (default):
https://github.com/user-attachments/assets/6b5de6ae-4fef-4c03-9835-1ce9a6c89239
| null | 9e933be762ada796e4d263739cda3360a3c38e91 | null | low | [
{
"filename": "packages/next/src/next-devtools/dev-overlay/components/overlay/styles.tsx",
"patch": "@@ -15,6 +15,8 @@ const styles = css`\n align-items: center;\n flex-direction: column;\n padding: 10vh 15px 0;\n+ /* color schemes we handle. Every other scheme the UA would need to overwrite ... |
vuejs/vue | 7,941 | feat(functional): add scopedSlots to context in functional components | Related #5381
We discussed this being added to 2.6.
It is nothing new, as `scopedSlots` already exist on `data`. The only reasons to add it are convenience and consistency with `slots` (although `slots` is a function).
If you have arguments for or against this feature, please share them 🙂
<!--
Please make su... | 5d52262f1ce56d080c3438c4773a81dc5c8397aa | fb6aa0609045e69a0b6050bc7b6466b63be8d69d | 1 | medium | [
{
"filename": "src/core/vdom/create-functional-component.js",
"patch": "@@ -48,6 +48,7 @@ export function FunctionalRenderContext (\n this.children = children\n this.parent = parent\n this.listeners = data.on || emptyObject\n+ this.scopedSlots = data.scopedSlots || emptyObject\n this.injections = r... |
nodejs/node | 61,350 | doc: restore @watilde to collaborators | Fixes: https://github.com/nodejs/TSC/issues/1813 | null | 1409ac4ffbc168359656dfbf23b923379976e59c | null | low | [
{
"filename": "README.md",
"patch": "@@ -453,6 +453,8 @@ For information about the governance of the Node.js project, see\n **Vladimir Morozov** <<vmorozov@microsoft.com>> (he/him)\n * [VoltrexKeyva](https://github.com/VoltrexKeyva) -\n **Mohammed Keyvanzadeh** <<mohammadkeyvanzade94@gmail.com>> (he/him... |
ollama/ollama | 9,742 | fix: error on embeddings; not currently implemented | embeddings is not (currently) supported in the ollama engine so error on requests to `/api/embed` for any models that are routed to that engine
without this change, embeddings on the ollama engine will return an empty list rather than error which hides the non-support
resolves #9739 | 65b0f329d1b6a9de6091c86a044b98fe6c543330 | ccfd41c4f0f1cbbb57ba19f79a2be36c07825b29 | 21 | medium | [
{
"filename": "runner/ollamarunner/runner.go",
"patch": "@@ -691,65 +691,6 @@ type EmbeddingResponse struct {\n \tEmbedding []float32 `json:\"embedding\"`\n }\n \n-func (s *Server) embeddings(w http.ResponseWriter, r *http.Request) {\n-\tvar req EmbeddingRequest\n-\tif err := json.NewDecoder(r.Body).Decode(... |
facebook/react | 33,572 | [compiler] Remove unnecessary fixture |
This is covered by iife-inline-ternary
---
[//]: # (BEGIN SAPLING FOOTER)
Stack created with [Sapling](https://sapling-scm.com). Best reviewed with [ReviewStack](https://reviewstack.dev/facebook/react/pull/33572).
* #33571
* #33558
* #33547
* #33543
* #33533
* #33532
* #33530
* #33526
* #33522
* #33518
* #33514
* #33... | 75e78d243f749d009fa1c5c09c3464301b992718 | ae962653d63ca88b6727d6f585026f2bbfa313a1 | 3 | medium | [
{
"filename": "compiler/packages/babel-plugin-react-compiler/src/__tests__/fixtures/compiler/error.todo-iife-inline-ternary.expect.md",
"patch": "@@ -1,33 +0,0 @@\n-\n-## Input\n-\n-```javascript\n-function Component(props) {\n- const x = props.foo\n- ? 1\n- : (() => {\n- throw new Error('Did ... |
vuejs/vue | 9,296 | [automated] Patreon sponsors update | This is an automated pull request. | 75a2b4cd42bcb864d37c3d4a12d984e746eb3a5f | cf42f8e870f34e59e19f9cbf7ac6309c3c9c160c | 1 | medium | [
{
"filename": "BACKERS.md",
"patch": "@@ -32,7 +32,7 @@ Funds donated via Patreon go directly to support Evan You's full-time work on Vu\n <tr>\n <td align=\"center\" valign=\"middle\">\n <a href=\"https://www.bitsrc.io/?utm_source=vue&utm_medium=vue&utm_campaign=vue&utm_term=vue&utm_conte... |
ggml-org/llama.cpp | 20,286 | Gracefully handle undetected tool parser, print error message. | Instead of crashing with an error when failing to detect proper toolcalling format in the autoparser, print an error message with instructions. | null | 1430c35948db08a35ba07ff3c2709bd50d58f563 | null | low | [
{
"filename": "common/chat-auto-parser-generator.cpp",
"patch": "@@ -3,6 +3,7 @@\n #include \"chat.h\"\n #include \"common.h\"\n #include \"json-schema-to-grammar.h\"\n+#include \"log.h\"\n #include \"nlohmann/json.hpp\"\n \n #include <stdexcept>\n@@ -182,7 +183,10 @@ common_peg_parser analyze_tools::build_... |
electron/electron | 49,547 | fix: chrome://accessibility drift | #### Description of Change
Refs CL:6870052
<details><summary>Before</summary>
<p>
<img width="793" height="596" alt="Screenshot 2026-01-27 at 11 34 10 AM" src="https://github.com/user-attachments/assets/56c8bd51-0ae9-4c24-b645-f17a270087b2" />
</p>
</details>
<details><summary>After</summary>
<p>
... | null | e6be04a7a0774eae527f5aba981c4ab7eaeabc74 | null | low | [
{
"filename": "shell/browser/ui/webui/accessibility_ui.cc",
"patch": "@@ -49,6 +49,10 @@\n #include \"ui/views/widget/widget.h\"\n #include \"ui/views/widget/widget_delegate.h\"\n \n+#if BUILDFLAG(IS_WIN)\n+#include \"ui/accessibility/platform/ax_platform_node_win.h\"\n+#endif\n+\n namespace {\n \n constexp... |
huggingface/transformers | 43,307 | [loading] Fix Transpose Operation, and qwen3_vl_moe mapping | # What does this PR do?
As per the title. Transpose needs to be more general in order to be used in reverse mode in a chain of several operations.
Supersedes https://github.com/huggingface/transformers/pull/43201 as it needed more work on the Transpose Operation!
Will also help https://github.com/huggingface/tra... | null | dfdede252984a1bf3a6e46aeb2a93ee75b4b51c6 | null | low | [
{
"filename": "src/transformers/conversion_mapping.py",
"patch": "@@ -82,6 +82,21 @@ def _build_checkpoint_conversion_mapping():\n operations=[MergeModulelist(dim=0)],\n ),\n ],\n+ \"qwen3_vl_moe\": [\n+ WeightConverter(\n+ source_patterns... |
huggingface/transformers | 43,258 | remove restriction for 8-bit bnb model when moving across devices | # What does this PR do?
As per title ! | 625ad844550caa3ff2f978827ad391a2e5ce6775 | 35a09896f997f0dac8b3f1c7b10e10b48321ae91 | 1 | medium | [
{
"filename": "src/transformers/modeling_utils.py",
"patch": "@@ -106,6 +106,7 @@\n copy_func,\n has_file,\n is_accelerate_available,\n+ is_bitsandbytes_available,\n is_env_variable_true,\n is_flash_attn_2_available,\n is_flash_attn_3_available,\n@@ -3529,10 +3530,9 @@ def to(self... |
ollama/ollama | 9,688 | ollama-debug.c: correct mistype | long long int -> long int | 9e4642e9b3e9a26d423c62915805375ca253d7d1 | 45a13b1dec1ccc1771cd28af826d603d9c885fd9 | 29 | medium | [
{
"filename": "ml/backend/ggml/ggml/src/ollama-debug.c",
"patch": "@@ -1,4 +1,5 @@\n #include <string.h>\n+#include <inttypes.h>\n \n #include \"ollama-debug.h\"\n \n@@ -24,7 +25,7 @@ static void print_tensor(const void *tensor, void (*cb)(const void *, int),\n fprintf(stderr, \"[\");\n for (int i =... |
nodejs/node | 61,348 | doc: run license-builder | License is likely out of date. This is an automatically generated PR by the `license-builder.yml` GitHub Action, which runs `license-builder.sh` and submits a new PR or updates an existing PR. | null | abfad122f808073fa325bcd61796b62bd0693c5a | null | low | [
{
"filename": "LICENSE",
"patch": "@@ -899,6 +899,32 @@ The externally maintained libraries used by Node.js are:\n written prior permission. M.I.T. makes no representations about the\n suitability of this software for any purpose. It is provided \"as is\"\n without express or implied warranty.... |
vercel/next.js | 88,845 | [ci] Fix Rspack/Turbopack test manifest generation | The recent change to always run all tests without aborting on failure (#88435) inadvertently broke manifest generation. Previously, test output was emitted for all tests when `NEXT_TEST_CONTINUE_ON_ERROR` was set, but that variable was removed. Now test output is only emitted for failing tests, causing the manifest to ... | null | 71f9bda3cda291eba9aece9ead1708d3c9de349b | null | low | [
{
"filename": ".github/workflows/integration_tests_reusable.yml",
"patch": "@@ -94,6 +94,7 @@ jobs:\n export NEXT_TEST_MODE=${{\n inputs.test_type == 'development' && 'dev' || 'start'\n }}\n+ export NEXT_TEST_EMIT_ALL_OUTPUT=1\n \n ${{ inputs.run_before_test }}\n \n@... |
ollama/ollama | 9,732 | sample: separate softmax and temperature transforms | Separating the transforms for readability | 4aeb67ef4c0c734e0077af39e0b8be1252662cc0 | 5c0b6639692ba5e6d44a0c73a9b5c85dc670d4f2 | 1 | medium | [
{
"filename": "sample/samplers.go",
"patch": "@@ -87,8 +87,8 @@ func (s *Sampler) sample(tokens []token) (token, error) {\n \t// topK also sorts the tokens in descending order of logits\n \ttokens = topK(tokens, s.topK)\n \n-\t// token logit values are updated to probabilities\n \ttokens = temperature(token... |
huggingface/transformers | 43,130 | [mimi] mimi conv cache edge case | # What does this PR do?
two things;
1. factorize the cache init so that it can be overwritten
2. handle the specific edge case with short hidden state: if lenght is < from padding, then the cache init should be taking into account in the padding states. This edge case is never matched with mimi but happens with ot... | 9eea1b0bd3b773b20610cceb61bd9e7a3ed61983 | 6d4d98410220a366c4bc78b175fac9db4d91c1f6 | 3 | medium | [
{
"filename": "src/transformers/models/kyutai_speech_to_text/modeling_kyutai_speech_to_text.py",
"patch": "@@ -141,18 +141,43 @@ def __init__(\n raise ValueError(\n f\"Expected `num_layers` ({num_layers}) values in `per_layer_padding`, `per_layer_padding_mode` and `per_layer_in_c... |
ollama/ollama | 9,669 | sample: fix sorting order and top_k | - the existing min heap implementation would fail for maintaining the correct sorting order - added container/heap implementation
- partial sorting instead of counting sort
- added tests to ensure floats were checked correctly
- [ ] vibe check the logits and model | 3ba91634c169ca27ff35f0c841d62ccae6403e47 | 4aeb67ef4c0c734e0077af39e0b8be1252662cc0 | 1 | medium | [
{
"filename": "sample/samplers.go",
"patch": "@@ -84,11 +84,8 @@ func (s *Sampler) sample(tokens []token) (token, error) {\n \t\treturn greedy(tokens), nil\n \t}\n \n-\tif s.topK > 0 {\n-\t\ttokens = topK(tokens, s.topK)\n-\t} else {\n-\t\tsortLogits(tokens)\n-\t}\n+\t// topK also sorts the tokens in descen... |
vercel/next.js | 88,850 | Turbopack: show compressed size for print_cache_item_size | ### What?
Improve the feature flag to show compressed size | 1bdb2e20052281c614a8bfef08a6ebb66a546129 | bed512a381250493a2e2e0f9185c69960c713fd9 | 3 | medium | [
{
"filename": "Cargo.lock",
"patch": "@@ -9431,6 +9431,7 @@ dependencies = [\n \"hashbrown 0.14.5\",\n \"indexmap 2.9.0\",\n \"lmdb-rkv\",\n+ \"lzzzz\",\n \"once_cell\",\n \"parking_lot\",\n \"rand 0.9.0\",",
"additions": 1,
"deletions": 0
},
{
"filename": "turbopack/crates/turbo-tasks... |
vuejs/vue | 9,128 | fix(ssr): renderToString cannot render comment | <!--
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 | b06c784b81a244e1bc2d028216fcd2ab873730b9 | null | low | [
{
"filename": "src/server/optimizing-compiler/codegen.js",
"patch": "@@ -225,7 +225,11 @@ function nodesToSegments (\n } else if (c.type === 2) {\n segments.push({ type: INTERPOLATION, value: c.expression })\n } else if (c.type === 3) {\n- segments.push({ type: RAW, value: escape(c.text) ... |
ggml-org/llama.cpp | 16,858 | CUDA: Remove unneded bias/gate dims in fused mmvq | Pointed out
[here](https://github.com/ggml-org/llama.cpp/pull/16847#discussion_r2476798989) that only a single value is needed per target col per thread | 8b11deea4663f29d3e042ce1056ba643264cd5f1 | d3dc9dd898be805c23a408cc36daed5b3bf29221 | 27 | medium | [
{
"filename": "ggml/src/ggml-cuda/mmvq.cu",
"patch": "@@ -190,26 +190,28 @@ static __global__ void mul_mat_vec_q(\n \n const uint32_t channel_bias = ids ? channel_x : channel_dst;\n \n- float x_biases[ncols_dst][rows_per_cuda_block] = { { 0.0f } };\n- float gate_biases[ncols_dst][rows_per_cuda_... |
facebook/react | 33,546 | [compiler] Add repro for IIFE in ternary causing a bailout | ---
[//]: # (BEGIN SAPLING FOOTER)
Stack created with [Sapling](https://sapling-scm.com). Best reviewed with [ReviewStack](https://reviewstack.dev/facebook/react/pull/33546).
* #33548
* __->__ #33546 | null | 75e78d243f749d009fa1c5c09c3464301b992718 | null | low | [
{
"filename": "compiler/packages/babel-plugin-react-compiler/src/__tests__/fixtures/compiler/error.todo-iife-inline-ternary.expect.md",
"patch": "@@ -0,0 +1,33 @@\n+\n+## Input\n+\n+```javascript\n+function Component(props) {\n+ const x = props.foo\n+ ? 1\n+ : (() => {\n+ throw new Error('Did ... |
electron/electron | 49,560 | fix: chrome://accessibility drift | Backport of #49547
See that PR for details.
Notes: Fixed an issue in `chrome://accessibility`. | null | d0661e765b698323bca09f86690519d17adaa33f | null | low | [
{
"filename": "shell/browser/ui/webui/accessibility_ui.cc",
"patch": "@@ -49,6 +49,10 @@\n #include \"ui/views/widget/widget.h\"\n #include \"ui/views/widget/widget_delegate.h\"\n \n+#if BUILDFLAG(IS_WIN)\n+#include \"ui/accessibility/platform/ax_platform_node_win.h\"\n+#endif\n+\n namespace {\n \n constexp... |
ollama/ollama | 9,692 | models/gemma3: remove final logit softcap | Softcap isn't in the whitepaper/implementation for the language model so we should remove it. There is no discernible difference in output with it removed.
With softcap:
```
❯ ollama run gemma3:27b
>>> I need to get warm. What button turns up the heat? /Users/bruce/Desktop/gemma-3-multimodal
... ity-example.png
... | a8e83a7654fffa169b90fa927e6d19c4c0c765d7 | a70820daa0a25024cfd857a528e717e3ac00a8e0 | 15 | medium | [
{
"filename": "model/models/gemma3/model_text.go",
"patch": "@@ -15,7 +15,6 @@ type TextOptions struct {\n \tattnKeyLen, attnValLen int\n \teps, ropeScale float32\n \tropeLocalBase, ropeGlobalBase float32\n-\tfinalLogitSoftcap float32\n \tlargeModelScaling ... |
nodejs/node | 61,345 | doc: clean up writing-and-running-benchmarks.md | doc: clean up writing and running benchmarks guide
**Summary of Changes:**
- Fixed spelling typo (prerequesites -> prerequisites).
- Wrapped bare shell commands in Markdown code blocks for better readability and copy-pasting. | null | b5bda89aa5ae94a783e1585ed3f4f0717e29ecc1 | null | low | [
{
"filename": "doc/contributing/writing-and-running-benchmarks.md",
"patch": "@@ -28,7 +28,7 @@ which need to be included in the global Windows `PATH`.\n \n If you are using Nix, all the required tools are already listed in the\n `benchmarkTools` argument of the `shell.nix` file, so you can skip those\n-pre... |
huggingface/transformers | 42,899 | Do not skip integration tests | # What does this PR do?
As per title, skipping tests caused regression after one of the huge refactors. Regression was fixed but we also need to enable the tests | 159219a6601ab87787b75cdc270b3b2438a9f0cf | 79f038631c5b2305b8b37dab4fc87ebb95c51cd6 | 25 | medium | [
{
"filename": "src/transformers/models/granitemoehybrid/modeling_granitemoehybrid.py",
"patch": "@@ -1482,8 +1482,8 @@ def forward(\n ```python\n >>> from transformers import AutoTokenizer, GraniteMoeHybridForCausalLM\n \n- >>> model = GraniteMoeHybridForCausalLM.from_pretrained(\"ibm... |
vercel/next.js | 88,854 | [test] Fix deploy test of `cache-components.server-action.test.ts` | When deployed, the sentinels where showing `at runtime` instead of `at buildtime`. There are two reasons for this:
1. The page sentinel was accidentally rendered in a client component, which causes a hydration mismatch, and the value toggling from `at buildtime` to `at runtime` when the page is hydrated before the a... | null | a3500c3227de0bda8e34c2a347047e88a524d9da | null | low | [
{
"filename": "test/e2e/app-dir/cache-components/app/server-action-inline/form.tsx",
"patch": "@@ -1,7 +1,6 @@\n 'use client'\n \n import { ReactNode, useActionState } from 'react'\n-import { getSentinelValue } from '../getSentinelValue'\n \n export function Form({ action }: { action: () => Promise<ReactNod... |
facebook/react | 33,527 | [ci] Don't skip experimental prerelease incorrectly |
Previously the experimental workflow relied on the canary one running first to avoid race conditions. However, I didn't account for the fact that the canary one can now be skipped.
| d60f77a533da830613431ddef83d0eda928697ad | 6b7e207cabe4c1bc9390d862dd9228e94e9edf4b | 1 | high | [
{
"filename": ".github/workflows/runtime_prereleases_manual.yml",
"patch": "@@ -88,6 +88,8 @@ jobs:\n # different versions of the same package, even if they use different\n # dist tags.\n needs: publish_prerelease_canary\n+ # Ensures the job runs even if canary is skipped\n+ if: always()\n... |
ggml-org/llama.cpp | 19,486 | CUDA: Update CCCL-tag for 3.2 to final release from RC | [CCCL 3.2 has been released](https://github.com/NVIDIA/cccl/releases/tag/v3.2.0
) since it was added to llama.cpp as part of the backend-sampling PR (#17004), and it makes sense to update from RC to final released version.
| null | 612db61886e0fd523f89a0f93be0ef19d450b7d7 | null | low | [
{
"filename": "ggml/src/ggml-cuda/CMakeLists.txt",
"patch": "@@ -64,7 +64,7 @@ if (CUDAToolkit_FOUND)\n FetchContent_Declare(\n CCCL\n GIT_REPOSITORY https://github.com/nvidia/cccl.git\n- GIT_TAG v3.2.0-rc2\n+ GIT_TAG v3.2.0\n G... |
vuejs/vue | 8,179 | Add iterable supports for v-for | <!--
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 | d40eb9c2880c8dd27fedb9fbc508823a15742274 | null | low | [
{
"filename": "src/core/instance/render-helpers/render-list.js",
"patch": "@@ -1,6 +1,6 @@\n /* @flow */\n \n-import { isObject, isDef } from 'core/util/index'\n+import { isObject, isDef, hasSymbol } from 'core/util/index'\n \n /**\n * Runtime helper for rendering v-for lists.\n@@ -25,11 +25,21 @@ export f... |
electron/electron | 49,561 | fix: chrome://accessibility drift | Backport of #49547
See that PR for details.
Notes: Fixed an issue in `chrome://accessibility`. | null | ad5c8483c770c75905cab0cabfce8ad5e612bae2 | null | low | [
{
"filename": "shell/browser/ui/webui/accessibility_ui.cc",
"patch": "@@ -49,6 +49,10 @@\n #include \"ui/views/widget/widget.h\"\n #include \"ui/views/widget/widget_delegate.h\"\n \n+#if BUILDFLAG(IS_WIN)\n+#include \"ui/accessibility/platform/ax_platform_node_win.h\"\n+#endif\n+\n namespace {\n \n constexp... |
ggml-org/llama.cpp | 18,964 | CUDA: Fix builds for older CCCL versions by ifdefing strided_iterator | Strided iterator was added in [CCCL 3.1](https://github.com/NVIDIA/cccl/releases/tag/v3.1.0), which is packaged into [CTK
13.1](https://docs.nvidia.com/cuda/cuda-toolkit-release-notes/index.html#id5)
Should fix #18960 | d1e3556481c8b351f9b7b69ba3febf6cb77fffa6 | 5bd341c9a135a13f901c4cacacc27fa5b299ce19 | 3 | high | [
{
"filename": "ggml/src/ggml-cuda/argsort.cu",
"patch": "@@ -2,6 +2,9 @@\n \n #ifdef GGML_CUDA_USE_CUB\n # include <cub/cub.cuh>\n+# if (CCCL_MAJOR_VERSION >= 3 && CCCL_MINOR_VERSION >= 1)\n+# define STRIDED_ITERATOR_AVAILABLE\n+# endif\n using namespace cub;\n #endif // GGML_CUDA_USE_CUB\n... |
ollama/ollama | 9,136 | cli: adding support ctrl-n/p like general cli | * adding ctrl-p like up
* adding ctrl-n like down | null | 6b45b1d6b4dff4e805d42c096c4db4e6d7be141e | null | low | [
{
"filename": "readline/readline.go",
"patch": "@@ -116,19 +116,9 @@ func (i *Instance) Readline() (string, error) {\n \n \t\t\tswitch r {\n \t\t\tcase KeyUp:\n-\t\t\t\tif i.History.Pos > 0 {\n-\t\t\t\t\tif i.History.Pos == i.History.Size() {\n-\t\t\t\t\t\tcurrentLineBuf = []rune(buf.String())\n-\t\t\t\t\t}... |
nodejs/node | 61,340 | tools: update nixpkgs-unstable to 3146c6aa9995e7351a398e17470e15305e6 | This is an automated update of nixpkgs-unstable to 3146c6aa9995e7351a398e17470e15305e6. | d050aa87e81973f2422eeb55d4926cf01def8020 | bb5d066989b4c45672832906d1bdac3d65d6b1e5 | 20 | medium | [
{
"filename": "tools/nix/pkgs.nix",
"patch": "@@ -1,10 +1,10 @@\n arg:\n let\n repo = \"https://github.com/NixOS/nixpkgs\";\n- rev = \"16c7794d0a28b5a37904d55bcca36003b9109aaa\";\n+ rev = \"3146c6aa9995e7351a398e17470e15305e6e18ff\";\n nixpkgs = import (builtins.fetchTarball {\n url = \"${repo}/ar... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.