repo stringclasses 15
values | fix_commit stringlengths 40 40 | buggy_commit stringlengths 40 40 | message stringlengths 3 64.3k | files listlengths 1 300 | timestamp timestamp[s]date 2013-03-13 20:45:00 2026-04-11 07:48:46 |
|---|---|---|---|---|---|
huggingface/transformers | be0b2d2a686cb0ad54185acb28166b0ec6fbca7c | a08936f6e45118b50f98278841cac39a26c2975d | Docs(it): fix typo in docstring wording (#44219)
Docs(it): fix docstring typo
Co-authored-by: matisgagneux21 <matisgagneux21@users.noreply.github.com> | [
{
"path": "docs/source/it/add_new_model.md",
"patch": "@@ -708,7 +708,7 @@ Siete quasi alla fine! L'ultima cosa rimasta é avere una bella docstring e una\n un template chiamato `docs/source/model_doc/brand_new_bert.rst`, che dovrete compilare. La prima cosa che un utente farà\n per usare il vostro modello s... | 2026-02-23T15:04:01 |
ollama/ollama | 92c2e8a56c7eb9a5a99439133220d707710da0f8 | 2e3fd86d482cb4e77e54179836ddd6a518e2300b | api: fix unreachable status err (#11423)
StatusError was unreachable, the client always checked for error messages in the response body first, and the server always includes error messages with HTTP error status codes. | [
{
"path": "api/client.go",
"patch": "@@ -222,10 +222,6 @@ func (c *Client) stream(ctx context.Context, method, path string, data any, fn f\n \t\t\treturn fmt.Errorf(\"unmarshal: %w\", err)\n \t\t}\n \n-\t\tif errorResponse.Error != \"\" {\n-\t\t\treturn errors.New(errorResponse.Error)\n-\t\t}\n-\n \t\tif re... | 2025-07-16T18:03:28 |
facebook/react | b160ee85223e80941c2361f28c417df349e7484e | 6d1d05ff91507ea844f332f65f5bc15cc8e10655 | [tests] Tests and patch for context variables
---
- [patch] Find context variables within FunctionDeclarations (previously
missing)
- [todo] Need to fix non-allocating values / variables being DCE'd
One approach is to label everything referenced by a lambda as context variables.
I couldn't come up with other ... | [
{
"path": "compiler/forget/packages/babel-plugin-react-forget/src/HIR/FindContextIdentifiers.ts",
"patch": "@@ -6,7 +6,9 @@ import { GeneratedSource } from \"./HIR\";\n type FindContextIdentifierState = {\n inLambda: number;\n currentLambda: Array<\n- NodePath<t.FunctionExpression> | NodePath<t.Arrow... | 2023-06-05T19:04:46 |
electron/electron | e1ed96b574a016c58b274ce0026432c05f5ad2aa | 9901d2f28173d9df4286bf4aa0f8c0726b184ec0 | test: scope internal test fixtures under `@electron-ci` (#33822)
* test: scope internal test fixtures under `@electron`
* Missed references
* Move packages from @electron to @electron-ci scope
* Fix tests | [
{
"path": "spec-main/fixtures/api/native-window-open-native-addon.html",
"patch": "@@ -7,7 +7,7 @@\n let requireError\n \n try {\n- echo = require('echo')\n+ echo = require('@electron-ci/echo')\n } catch (error) {\n requireError = error\n }",
"additions": 1,
"deletions": 1,
"la... | 2022-05-02T17:09:23 |
golang/go | 31e50af5f3366a9fc4e91a589f91ad579b7cba56 | b47ce8b0e997f09abd6e1e5c23eb59d6e9d5be2c | crypto/rsa: revert minimum GenerateKey size to 32 bits
No point in causing breakage even with GODEBUG=rsa1024min=0.
Change-Id: I923254a8c8afaca77be551b19e3555c44ebdbb67
Reviewed-on: https://go-review.googlesource.com/c/go/+/636557
Reviewed-by: Roland Shoemaker <roland@golang.org>
Reviewed-by: David Chase <drchase@goo... | [
{
"path": "src/crypto/internal/fips140/rsa/keygen.go",
"patch": "@@ -13,9 +13,9 @@ import (\n )\n \n // GenerateKey generates a new RSA key pair of the given bit size.\n-// bits must be at least 128.\n+// bits must be at least 32.\n func GenerateKey(rand io.Reader, bits int) (*PrivateKey, error) {\n-\tif bi... | 2024-12-16T16:33:50 |
huggingface/transformers | a08936f6e45118b50f98278841cac39a26c2975d | 9140fbee216c40fb18141e5a2ed24d5a96782dcc | fix bug with position_ids on qwen3-vl models, such that position_ids include text position (#44158)
* fix derived text position
* add tests for position ids in qwen3-vl
* fix ruff error
---------
Co-authored-by: Raushan Turganbay <raushan@huggingface.co> | [
{
"path": "src/transformers/models/qwen3_5/modeling_qwen3_5.py",
"patch": "@@ -1339,17 +1339,17 @@ def forward(\n past_seen_tokens, past_seen_tokens + inputs_embeds.shape[1], device=inputs_embeds.device\n )\n \n- # mrope: the hard coded `3` is for temporal, height and widt... | 2026-02-23T14:53:33 |
vercel/next.js | 5d85a444c4ee22d7611e20e4adaba324d34d88eb | 245406adf09562ec27f21aaf7a867e9b745741b9 | fix(CI): Avoid rebuilding native code in jobs that don't depend on native code (#78944)
The following jobs do not depend on `build-native`, but don't specify `skipNativeBuild`:
- `lint`
- `test-devlow`
- `test-next-swc-wasm`
- `test-next-swc-napi-wasi` *(currently disabled)*
- `test-unit` *(does actually appear to de... | [
{
"path": ".github/workflows/build_and_test.yml",
"patch": "@@ -86,6 +86,8 @@ jobs:\n needs: ['build-next']\n uses: ./.github/workflows/build_reusable.yml\n with:\n+ skipNativeBuild: 'yes'\n+ skipNativeInstall: 'yes'\n afterBuild: |\n pnpm lint-no-typescript\n pnp... | 2025-05-08T08:31:01 |
facebook/react | 5ba0a03e00f3bc1f1225989cd168d037d74a194f | a65e9cd1f13aed33c12c5ea20ee21cacd414ac22 | [config] Clean up codegen imports and gating
- Gating checks for debugging / profiling can be moved to within the
instrumentation or makeReadOnly functions. This simplifies codegen and reduces
code bloat.
- Warn if importing conflicting identifiers
- Aggregate imports from the same source | [
{
"path": "compiler/forget/packages/babel-plugin-react-forget/src/Entrypoint/Instrumentation.ts",
"patch": "@@ -11,18 +11,13 @@ import * as t from \"@babel/types\";\n export function addInstrumentForget(\n fn: NodePath<t.FunctionDeclaration>,\n fnName: string,\n- gatingIdentifierName: string,\n instr... | 2023-06-05T18:46:59 |
ollama/ollama | 2e3fd86d482cb4e77e54179836ddd6a518e2300b | 4261a3b0b264430489921a1b4a16a6267711d595 | docs: fix typo in macos.md (#11425) | [
{
"path": "docs/macos.md",
"patch": "@@ -22,7 +22,7 @@ To install the Ollama application somewhere other than `Applications`, place the\n Ollama on MacOS stores files in a few different locations.\n - `~/.ollama` contains models and configuration\n - `~/.ollama/logs` contains logs\n- - *app.log* contains... | 2025-07-16T17:50:46 |
electron/electron | 9901d2f28173d9df4286bf4aa0f8c0726b184ec0 | 14f07d7814e3ffd64d18213129ff4a4cfa50e1e1 | fix: support mixed-case extensions in Linux file dialogs (#33918) | [
{
"path": "shell/browser/ui/file_dialog_gtk.cc",
"patch": "@@ -31,6 +31,25 @@ namespace {\n static const int kPreviewWidth = 256;\n static const int kPreviewHeight = 512;\n \n+std::string MakeCaseInsensitivePattern(const std::string& extension) {\n+ std::string pattern(\"*.\");\n+\n+ for (std::size_t i = ... | 2022-05-02T15:54:17 |
golang/go | dd7a7ba38f36dd6abc1e14b8d0e8bf05a5383161 | 427a2401af088716bf5b394156e45fb4de54dc5a | crypto/internal/fips140/aes: mark AES-ECB as not approved
NIST SP 800-131Ar3 ipd, scheduled for publication in 2025Q1, marks
AES-ECB as disallowed for encryption, and legacy use for decryption.
There are apparently no details on how the transition is going to work,
so to avoid surprises we just mark direct use of the... | [
{
"path": "src/crypto/internal/fips140/aes/aes.go",
"patch": "@@ -94,6 +94,8 @@ func newBlockExpanded(c *blockExpanded, key []byte) {\n func (c *Block) BlockSize() int { return BlockSize }\n \n func (c *Block) Encrypt(dst, src []byte) {\n+\t// AES-ECB is not approved in FIPS 140-3 mode.\n+\tfips140.RecordNo... | 2024-12-16T18:24:20 |
huggingface/transformers | 65935e09212bcd5c3ceecbd6e12ace3ec26a804f | 9f43e7b4e2e92044e56f4c68a60c17da1a514fed | fix(testing): Fix LayoutXLM tokenization test and LightOnOCR SDPA flash test failures on main CI (#43988)
fix: Fix LayoutXLM and LightOnOCR test failures on main
Co-authored-by: Ita Zaporozhets <31893021+itazap@users.noreply.github.com> | [
{
"path": "tests/test_modeling_common.py",
"patch": "@@ -3599,6 +3599,7 @@ def test_sdpa_can_dispatch_on_flash(self):\n \"jamba\",\n \"kosmos-2\",\n \"mllama\",\n+ \"lighton_ocr\",\n \"pixtral\",\n \"sam\",\n ... | 2026-02-23T13:59:50 |
facebook/react | a65e9cd1f13aed33c12c5ea20ee21cacd414ac22 | 40d1459115f53c4a94201abebe082de1ad2e39ec | [devx] emit calls to `makeReadOnly` for debugging
---
Emit calls to makeReadOnly for memoized values.
```js
function MyComponent() {
let x;
if (c_0) {
x = // ... (recompute x)
$[0] = __DEV__ ? makeReadOnly(x, "MyComponent") : x;
} else {
x = $[0]
}
}
```
- import source / specifier should be... | [
{
"path": "compiler/forget/packages/babel-plugin-react-forget/src/Entrypoint/Program.ts",
"patch": "@@ -12,10 +12,10 @@ import {\n CompilerErrorDetail,\n ErrorSeverity,\n } from \"../CompilerError\";\n-import { compileFn } from \"./Pipeline\";\n import { GeneratedSource } from \"../HIR\";\n import { add... | 2023-06-05T18:46:58 |
ollama/ollama | 35fda7b4af556e7eeef2b5dcb3638435382b2576 | 66fb8575ced090a969c9529c88ee57a8df1259c2 | ggml: Report ordinal IDs for AMD GPUs on Windows
We don't get valid UUIDs for AMD GPUs on Windows, so the best option
is to use the ordinal IDs. This brings us in line with what we currently
do on the Ollama server - the only exception is AMD GPUs on Linux, which
falls back to using ordinal IDs. The GGML implementatio... | [
{
"path": "llama/patches/0017-ggml-Export-GPU-UUIDs.patch",
"patch": "@@ -7,41 +7,41 @@ This enables matching up devices and information reported by the backend\n with tools (e.g. nvidia-smi) and system management libraries (e.g. nvml).\n ---\n ggml/include/ggml-backend.h | 1 +\n- ggml/src/ggml-cuda/... | 2025-06-26T00:13:32 |
golang/go | 427a2401af088716bf5b394156e45fb4de54dc5a | 75736cc169631dfaaa48ad46a00e1bf31d830ec8 | cmd/go/testdata/script: update test_flags for new test output
With CL 635856, an error for a package missing go files is now caught at
load time rather than at build time, so it's reported differently.
Update the test to check for the correct output unbreak the builders.
For #70820
Change-Id: I91fcac17961635b9a92782... | [
{
"path": "src/cmd/go/testdata/script/test_flags.txt",
"patch": "@@ -15,8 +15,8 @@ stdout '\\Aok\\s+example.com/x\\s+[0-9.s]+\\n\\z'\n # Even though ./x looks like a package path, the real package should be\n # the implicit '.'.\n ! go test --answer=42 ./x\n-stdout '^FAIL\\t. \\[build failed\\]'\n-stderr '^... | 2024-12-13T22:33:16 |
electron/electron | 14f07d7814e3ffd64d18213129ff4a4cfa50e1e1 | 6f851afab53e17df4b9c76cbcf82ef2a51f4d040 | fix: move FirstPartySets into the browser process (#33998)
Ref: https://chromium-review.googlesource.com/c/chromium/src/+/3448551 | [
{
"path": "shell/browser/electron_browser_main_parts.cc",
"patch": "@@ -25,6 +25,7 @@\n #include \"content/public/browser/browser_thread.h\"\n #include \"content/public/browser/child_process_security_policy.h\"\n #include \"content/public/browser/device_service.h\"\n+#include \"content/public/browser/first_... | 2022-05-02T15:20:30 |
facebook/react | 3744930728bfe6a08c382b25834e258bc44bc07f | 1e0eb25cd07ef8a1aef6cf6acba04e2a0439e834 | Fix last(?) order of evaluation bug
Not to get ahead of myself (sorry i had to), but i think this is the last order
of evaluation bug. At least it's the last one we know of[1]. Per the previous
PR, the issue is that constant propagation can copy the last value of a sequence
expression to where the sequence is used... | [
{
"path": "compiler/forget/src/HIR/BuildHIR.ts",
"patch": "@@ -1171,7 +1171,7 @@ function lowerExpression(\n const continuationBlock = builder.reserve(builder.currentBlockKind());\n const place = buildTemporaryPlace(builder, exprLoc);\n \n- const sequenceBlock = builder.enter(\"value\", (_)... | 2023-06-05T01:41:27 |
ggml-org/llama.cpp | 8a65a7a8ee70061ba180b6634157a8323f293087 | 8a132faaa06d904bbeb3ebfbc8f7e2ed921b8ded | ci: drop v5 `all:` composition from labeler.yml (#21627)
actions/labeler@v6 removed the `all:` / `any:` composition keys.
The `server/webui` and `server` entries used `all:` to combine
`any-glob-to-any-file` with negated `all-globs-to-all-files`,
which now errors on every PR with:
Unknown config options were unde... | [
{
"path": ".github/labeler.yml",
"patch": "@@ -75,21 +75,13 @@ android:\n - examples/llama.android/**\n server/webui:\n - changed-files:\n- - all:\n- - any-glob-to-any-file:\n- - tools/server/webui/**\n- - tools/server/public/**\n- -... | 2026-04-09T06:20:19 |
huggingface/transformers | 9f43e7b4e2e92044e56f4c68a60c17da1a514fed | 78c873efeef8aec243568c4536175b052353f024 | Add ModernVBERT models (#42504)
* modernbert modeling
* init back
* add mdoernvbert
* refine test: testing logits
* refine test: testing logits
* add missing methods to modernvbert + clean code and doc
* add modernvbert and colmodernvbert
* clean
* add auto mapping for colmvb
* colmvb imports
* add colmvb te... | [
{
"path": "docs/source/en/_toctree.yml",
"patch": "@@ -671,6 +671,8 @@\n title: ModernBert\n - local: model_doc/modernbert-decoder\n title: ModernBERTDecoder\n+ - local: model_doc/modernvbert\n+ title: ModernVBert\n - local: model_doc/mpnet\n title: MPNet\n ... | 2026-02-23T12:55:41 |
golang/go | 75736cc169631dfaaa48ad46a00e1bf31d830ec8 | 1218566fe5c90e2bd39693566453828e6fbb4c12 | fmt, strconv: document that exponent is always two digits
Except for %b where it is only one.
Fixes #70862
Change-Id: Ic423a799b73bfa534f4083f6544bb9cd639fef06
Reviewed-on: https://go-review.googlesource.com/c/go/+/636915
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Revie... | [
{
"path": "src/fmt/doc.go",
"patch": "@@ -50,6 +50,9 @@ Floating-point and complex constituents:\n \t%x\thexadecimal notation (with decimal power of two exponent), e.g. -0x1.23abcp+20\n \t%X\tupper-case hexadecimal notation, e.g. -0X1.23ABCP+20\n \n+\tThe exponent is always a decimal integer.\n+\tFor format... | 2024-12-16T23:38:43 |
electron/electron | 015185ad5aa54721e434ba0cd9296ffa8642c680 | fb534c927ae3ca696f0de3a10f90d447dd4fd52c | fix: fix FirstPartySetsHandler initialization (#33971) | [
{
"path": "shell/browser/net/system_network_context_manager.cc",
"patch": "@@ -20,6 +20,7 @@\n #include \"components/os_crypt/os_crypt.h\"\n #include \"components/prefs/pref_service.h\"\n #include \"content/public/browser/browser_thread.h\"\n+#include \"content/public/browser/first_party_sets_handler.h\"\n ... | 2022-04-29T05:17:19 |
facebook/react | 1e0eb25cd07ef8a1aef6cf6acba04e2a0439e834 | cd115ec128a80a69c4aac253c853d12026f1ea61 | Use sequence terminal, fix most remaining order-of-evaluation bugs
Changes the lowering for sequence expressions to use the new terminal. When
converting to a ReactiveFunction, we convert these terminals into
ReactiveSequenceValues, which nests the instructions and preserves order of
evaluation in the output.
Th... | [
{
"path": "compiler/forget/src/HIR/BuildHIR.ts",
"patch": "@@ -1168,23 +1168,48 @@ function lowerExpression(\n const expr = exprPath as NodePath<t.SequenceExpression>;\n const exprLoc = expr.node.loc ?? GeneratedSource;\n \n- let last: Place | null = null;\n- for (const item of expr.ge... | 2023-06-05T01:26:01 |
ggml-org/llama.cpp | 2dcb7f74edde532831c5ee9a9b6c403f2b68854f | 660600081fb7b9b769ded5c805a2d39a419f0a0d | fix: free ctx_copy in ggml_opt_free to plug per-training-session leak (#21592)
* fix: free ctx_copy in ggml_opt_free to plug per-training-session leak
ggml_opt_alloc populates opt_ctx->ctx_copy via a free+init pair every
time the allocated graph shape changes. The last ctx_copy from the
final ggml_opt_alloc call surv... | [
{
"path": "ggml/src/ggml-opt.cpp",
"patch": "@@ -589,6 +589,7 @@ void ggml_opt_free(ggml_opt_context_t opt_ctx) {\n ggml_backend_buffer_free(opt_ctx->buf_cpu);\n ggml_free(opt_ctx->ctx_static);\n ggml_free(opt_ctx->ctx_cpu);\n+ ggml_free(opt_ctx->ctx_copy);\n delete opt_ctx;\n }\n ",
... | 2026-04-08T15:40:15 |
huggingface/transformers | 78c873efeef8aec243568c4536175b052353f024 | a3dcad9e25af4c8747a365ec3f9a6b33e4b9abc1 | fix: HiggsAudioV2 cached decode inputs in compiled generation (#44201)
* Fix HiggsAudioV2 cached decode inputs in compiled generation
* tweaked fix
* missing block
---------
Co-authored-by: Cyril Vallez <cyril.vallez@gmail.com> | [
{
"path": "src/transformers/models/higgs_audio_v2/generation_higgs_audio_v2.py",
"patch": "@@ -265,18 +265,21 @@ def _sample(\n else self.__call__\n )\n \n- # Assisted generation completes the prefill stage in candidate generator so that\n- # we don't have several `prefill`... | 2026-02-23T12:39:17 |
golang/go | d92c34a3870bace34724e69ec2516d59ae432d32 | 3bd08b97921826c1b0a5fbf0789f4b49d7619977 | cmd/go: don't create test actions for incomplete packages
If a package is incomplete, don't create the actions for building and
testing it. Instead report the errors for the package's dependencies
and report a setup failed error (similar to what we'd to for a load
error when producing the test packages). This produces... | [
{
"path": "src/cmd/go/internal/load/pkg.go",
"patch": "@@ -3068,7 +3068,15 @@ func setPGOProfilePath(pkgs []*Package) {\n // CheckPackageErrors prints errors encountered loading pkgs and their\n // dependencies, then exits with a non-zero status if any errors were found.\n func CheckPackageErrors(pkgs []*Pa... | 2024-12-13T22:33:16 |
facebook/react | cd115ec128a80a69c4aac253c853d12026f1ea61 | e410815aeb10bbef6406e09cf0783187fae34d2f | Scaffold for sequence terminal
I realized that we can use our value block system to fix _most_ of the remaining
order-of-evaluation issues we had with sequence expressions. This PR adds a new
SequenceTerminal to HIR; there is already a ReactiveFunction equivalent
(ReactiveSequenceValue) that the next PR will conve... | [
{
"path": "compiler/forget/src/HIR/HIR.ts",
"patch": "@@ -279,7 +279,8 @@ export type Terminal =\n | LogicalTerminal\n | TernaryTerminal\n | OptionalTerminal\n- | LabelTerminal;\n+ | LabelTerminal\n+ | SequenceTerminal;\n \n function _staticInvariantTerminalHasLocation(\n terminal: Terminal\n@@ -... | 2023-06-05T01:25:59 |
ggml-org/llama.cpp | e9fd96283d75b55f46f198a68507825f400127d4 | 3ba12fed0a50af94bd9cfdea6f0b59e5aba8ed4a | Propose fix a couple of typos (#21581)
Signed-off-by: John E <jeis4wpi@outlook.com> | [
{
"path": "tests/test-chat.cpp",
"patch": "@@ -3454,7 +3454,7 @@ static void test_template_output_peg_parsers(bool detailed_debug) {\n },\n \"replaceAll\": {\n \"type\": \"boolean\",\n- \"d... | 2026-04-08T14:29:03 |
electron/electron | fb534c927ae3ca696f0de3a10f90d447dd4fd52c | 192a7fad0d548d1883c58bdf95ab7a2ff1391881 | refactor: better solution for resizable frameless DCHECK (#33790)
* refactor: better solution for resizable frameless DCHECK
* fix: also implement TargetForRectin WinFrameView | [
{
"path": "shell/browser/ui/views/frameless_view.cc",
"patch": "@@ -112,6 +112,16 @@ void FramelessView::UpdateWindowTitle() {}\n \n void FramelessView::SizeConstraintsChanged() {}\n \n+views::View* FramelessView::TargetForRect(views::View* root,\n+ const gfx::Rect& ... | 2022-04-29T00:34:12 |
huggingface/transformers | a3dcad9e25af4c8747a365ec3f9a6b33e4b9abc1 | 9f7faed251b433c999b3d8f4313e675b7071165b | fix(models): Fix LayoutLMv2 NER crash and broken batched truncation/padding (#44187)
fix: Add only_label_first_subword and model_max_length | [
{
"path": "src/transformers/models/layoutlmv2/tokenization_layoutlmv2.py",
"patch": "@@ -175,6 +175,7 @@ def __init__(\n only_label_first_subword=True,\n tokenize_chinese_chars=True,\n strip_accents=None,\n+ model_max_length=512,\n **kwargs,\n ):\n self.do_... | 2026-02-23T10:29:37 |
golang/go | 3bd08b97921826c1b0a5fbf0789f4b49d7619977 | 18b5435fc84225ca303da7a110c7e8065dc4bbda | runtime: usleep in TestWeakToStrongMarkTermination
There's a subtle bug in this test (big surprise): time.Sleep allocates,
so the time.Sleep(100*time.Millisecond) before unblocking gcMarkDone
might itself end up in gcMarkDone.
Work around this by using usleep here instead.
Fixes #70532.
Change-Id: I4c642ebb12f737cd... | [
{
"path": "src/runtime/gc_test.go",
"patch": "@@ -834,7 +834,11 @@ func TestWeakToStrongMarkTermination(t *testing.T) {\n \t\tdone <- struct{}{}\n \t}()\n \tgo func() {\n-\t\ttime.Sleep(100 * time.Millisecond)\n+\t\t// Usleep here instead of time.Sleep. time.Sleep\n+\t\t// can allocate, and if we get unluck... | 2024-12-13T23:18:59 |
ggml-org/llama.cpp | 87f4744a805f7837b95ce7ff94fc60eb7a3fbcc8 | 85d482e6b6706648070f620797e54f1a6a0ff3d8 | examples : disable cb_eval callback for --save-logits (#21553)
This commit updates the debug example to not create the
base_callback_data.
The motivation for this is when using `--save-logits`, which is used by
examples/model-conversion scripts, we often don't care about the tensor
outputs and they just add noise to ... | [
{
"path": "examples/debug/debug.cpp",
"patch": "@@ -9,6 +9,7 @@\n #include <vector>\n #include <filesystem>\n #include <fstream>\n+#include <optional>\n #include <regex>\n \n static void print_usage(int /*argc*/, char ** argv) {\n@@ -222,7 +223,10 @@ int main(int argc, char ** argv) {\n llama_backend_in... | 2026-04-08T12:10:33 |
facebook/react | e410815aeb10bbef6406e09cf0783187fae34d2f | 084edadaa1fe173d2c89e022c570e0b1bcdc405e | Infer more primitive types
Handles three more cases:
* Template literals
* deletion (property/computed)
* type casts
Only the latter has an observable impact, though i added tests for deletion just
in case and found a bug. For type casts, they're reasonably common internally
for fixmes, so this PR will help... | [
{
"path": "compiler/forget/src/Inference/InferReferenceEffects.ts",
"patch": "@@ -837,7 +837,8 @@ function inferBlock(\n state.reference(instrValue.object, Effect.Mutate);\n state.reference(instrValue.property, Effect.Read);\n state.initialize(instrValue, ValueKind.Immutable);\n- ... | 2023-06-05T00:27:24 |
golang/go | 18b5435fc84225ca303da7a110c7e8065dc4bbda | c1f2542c8bef27872a95efd7904afeeee36fe976 | testing: don't measure cleanup time after B.Loop
B.Loop resets the timer on the first iteration so that setup code
isn't measured, but it currently leaves the timer running after the
last iteration, meaning that cleanup code will still be measured. Fix
this by stopping the timer when B.Loop returns false to indicate t... | [
{
"path": "src/testing/benchmark.go",
"patch": "@@ -366,6 +366,8 @@ func (b *B) ReportMetric(n float64, unit string) {\n func (b *B) stopOrScaleBLoop() bool {\n \ttimeElapsed := highPrecisionTimeSince(b.start)\n \tif timeElapsed >= b.benchTime.d {\n+\t\t// Stop the timer so we don't count cleanup time\n+\t\... | 2024-12-13T02:18:44 |
electron/electron | 192a7fad0d548d1883c58bdf95ab7a2ff1391881 | 097da1d4ba79a78215eb324a23c0ac0cfcae29d5 | fix: disable MallocNanoZone on mac (#33704) | [
{
"path": "shell/browser/resources/mac/Info.plist",
"patch": "@@ -41,6 +41,11 @@\n <false/>\n <key>NSQuitAlwaysKeepsWindows</key>\n <false/>\n+ <key>LSEnvironment</key>\n+ <dict>\n+ <key>MallocNanoZone</key>\n+ <string>0</string>\n+ </dict>\n <key>NSMicrophoneUsageDescript... | 2022-04-28T14:28:27 |
huggingface/transformers | 9f7faed251b433c999b3d8f4313e675b7071165b | 5ef42f53c4225d9af28e8a297afcc624b8291a06 | [CB] [Major] Asynchronous batching (#43960)
* Cleanup: batch is more self contained
* Created utils.py file
* Moved pad to utils
* Pin memory for input and outputs
* Consolidate inputs into a bulk tensor
* Consolidated read and write indices
* Add the transfer_inputs fn
* Renames and getters
* Remove useless s... | [
{
"path": "benchmark_v2/benchmark_scripts/continuous_batching_overall.py",
"patch": "@@ -54,6 +54,8 @@ def run_and_parse_cb_example(args: str) -> dict:\n \n # Benchmark with high number of samples\n results.append(run_and_parse_cb_example(\"--samples 500\"))\n+ # Benchmark with async API\n+ re... | 2026-02-23T10:11:26 |
ollama/ollama | 44b17d2bfa0073e012679152421c0b69671d380e | 3b8b692218bf0da28859c89b71f4f5731f29002c | tools: fix parsing tool calls with empty arguments, missing required fields (#11233) | [
{
"path": "tools/tools.go",
"patch": "@@ -134,16 +134,16 @@ func (p *Parser) parseToolCall() *api.ToolCall {\n \t\treturn nil\n \t}\n \n-\t// only look for arguments if the tool has parameters\n+\t// only look for arguments after the tool name if the tool has parameters\n+\t// TODO (jmorganca): while probab... | 2025-06-30T15:59:03 |
ggml-org/llama.cpp | 85d482e6b6706648070f620797e54f1a6a0ff3d8 | ae65fbdf33b4ccc7a5c8fb4815fe86c1e7b278f5 | parser: fix MiniMax handling (#21573) | [
{
"path": "common/chat.cpp",
"patch": "@@ -1963,7 +1963,7 @@ static common_chat_params common_chat_templates_apply_jinja(const struct common_\n params.add_generation_prompt = true;\n std::string gen_prompt = common_chat_template_direct_apply_impl(tmpl, params);\n auto diff ... | 2026-04-08T10:47:25 |
golang/go | 090748d6c7973e9bb8f5efe069135c8ea0f0d89c | e39e965e0e0cce65ca977fd0da35f5bfb68dc2b8 | testing: improve B.Loop docs, use B.Loop in examples
This updates the testing documentation to frame B.Loop as the
canonical way to write benchmarks. We retain documentation on b.N
benchmarks because people will definitely continue to see them (and
write them), but it's demoted to clearly second class.
This also atte... | [
{
"path": "src/testing/benchmark.go",
"patch": "@@ -78,7 +78,7 @@ type InternalBenchmark struct {\n }\n \n // B is a type passed to [Benchmark] functions to manage benchmark\n-// timing and to specify the number of iterations to run.\n+// timing and control the number of iterations.\n //\n // A benchmark en... | 2024-12-12T20:43:37 |
facebook/react | 084edadaa1fe173d2c89e022c570e0b1bcdc405e | e41a8d6a1ea481be0fd1b4b3cbe4930919525546 | Distinguish error fixtures for invalid code
Renames some error fixtures for clarity, "error.invalid-*" are fixtures that are
expected to fail for invalid input, where other "error.*" fixtures are basically
todos. While i was here i clarified the error messages for invalid useMemo
callbacks, and changes the error s... | [
{
"path": "compiler/forget/src/Inference/InlineUseMemo.ts",
"patch": "@@ -100,15 +100,15 @@ export function inlineUseMemo(fn: HIRFunction): void {\n }\n \n if (body.loweredFunc.params.length > 0) {\n- CompilerError.invariant(\n- \"Did not expect any argume... | 2023-06-04T18:41:42 |
electron/electron | b5297ea8e2380b9174c26d5a35b1a73c2c9566d0 | 15c931201ab880de84009baa59b9bc5923e1ef67 | docs: fix return type of setJumpList() in electron.d.ts (#33910)
Before:
```ts
setJumpList(categories: (JumpListCategory[]) | (null)): void;
```
After:
```ts
setJumpList(categories: (JumpListCategory[]) | (null)): ('ok' | 'error' | 'invalidSeparatorError' | 'fileTypeRegistrationError' | 'customCategoryAccessDe... | [
{
"path": "docs/api/app.md",
"patch": "@@ -864,6 +864,8 @@ Returns `Object`:\n \n * `categories` [JumpListCategory[]](structures/jump-list-category.md) | `null` - Array of `JumpListCategory` objects.\n \n+Returns `string`\n+\n Sets or removes a custom Jump List for the application, and returns one of the\n ... | 2022-04-28T10:15:23 |
ollama/ollama | 45f216a9c7e65bd30ab0e2b1b9fdb7cb2ad9436d | d0b32def60b413407ddf4b4b063ba105a1ef2f92 | ggml: Temporarily disable reporting UUIDs
This is causing segfaults, so disable it. Currently UUIDs are only
used for debugging purposes, although they planned to be used in
additional ways in the future.
Bug #11211 | [
{
"path": "ml/backend/ggml/ggml.go",
"patch": "@@ -138,7 +138,10 @@ func New(modelPath string, params ml.BackendParams) (ml.Backend, error) {\n \trequiredMemory.CPU.Name = C.GoString(C.ggml_backend_dev_name(cpuDeviceBufferType.d))\n \tvar props C.struct_ggml_backend_dev_props\n \tC.ggml_backend_dev_get_prop... | 2025-06-27T18:11:49 |
huggingface/transformers | 5ef42f53c4225d9af28e8a297afcc624b8291a06 | df1cd3a7bfe3e91a4c9bc32cd273d3683f976da6 | Fix LASR feature extractor regression from invalid center argument (#44207)
* Fix LASR feature extractor center arg regression
* Keep LASR center API while fixing extractor mismatch
* revert lasr feature extractor
Co-authored-by: nightcityblade <nightcityblade@users.noreply.github.com>
* voxtral realtime not modul... | [
{
"path": "src/transformers/models/lasr/feature_extraction_lasr.py",
"patch": "@@ -149,7 +149,6 @@ def __call__(\n do_normalize: bool | None = None,\n device: str | None = \"cpu\",\n return_token_timestamps: bool | None = None,\n- center: bool = True,\n **kwargs,\n ... | 2026-02-23T10:01:35 |
golang/go | 08770a5b944ba1cf4f62f075bfd94cd36a061bdb | c4f356dd863c449835248c24f6dc653323010a1c | cmd/link: make dwarf name slice index self-describing
cmd/compile/internal/dwarfgen.createComplexVar does it this way, which
has the nice property of documenting the expected prefix.
This is primarily for newtype, since defgotype checks for the prefix
immediately prior, but I changed both for consistency.
Change-Id:... | [
{
"path": "src/cmd/link/internal/ld/dwarf.go",
"patch": "@@ -520,7 +520,7 @@ func (d *dwctxt) defgotype(gotype loader.Sym) loader.Sym {\n \t\td.linkctxt.Errorf(gotype, \"dwarf: type name doesn't start with \\\"type:\\\"\")\n \t\treturn d.mustFind(\"<unspecified>\")\n \t}\n-\tname := sn[5:] // could also dec... | 2024-12-13T20:05:27 |
electron/electron | 15c931201ab880de84009baa59b9bc5923e1ef67 | f91b24deb5d573563b0fd904ea24cbf078d7660e | chore: remove git.io (#33933)
* chore: remove git.io
All links on git.io will stop redirecting after April 29, 2022.
- https://github.blog/changelog/2022-04-25-git-io-deprecation/
* fix: lint
Co-authored-by: Charles Kerr <charles@charleskerr.com>
Co-authored-by: Charles Kerr <charles@charleskerr.com> | [
{
"path": "shell/browser/native_window_mac.mm",
"patch": "@@ -364,7 +364,8 @@ void ViewDidMoveToSuperview(NSView* self, SEL _cmd) {\n // Don't show title bar.\n [window_ setTitlebarAppearsTransparent:YES];\n [window_ setTitleVisibility:NSWindowTitleHidden];\n- // Remove non-transparent corner... | 2022-04-27T19:56:01 |
ollama/ollama | ad118d8b1306063d0ab2e967a20bd7914a135b28 | f08534137b127e1159f4b162881b8c2defc93158 | ci: arm sbsa fixes (#11194) | [
{
"path": ".github/workflows/release.yaml",
"patch": "@@ -322,12 +322,12 @@ jobs:\n - run: |\n for COMPONENT in bin/* lib/ollama/*; do\n case \"$COMPONENT\" in\n- bin/ollama) echo $COMPONENT >>ollama-${{ matrix.os }}-${{ matrix.arch }}.tar.in ;;\n- ... | 2025-06-25T04:00:15 |
ggml-org/llama.cpp | 09343c01986667b26d414733950c9d38e91c3065 | 97508acb17ff933c67edb3a2f0997bc19a6abb98 | model : support step3-vl-10b (#21287)
* feat: support step3-vl-10b
* use fused QKV && mapping tensor in tensor_mapping.py
* guard hardcoded params and drop crop metadata
* get understand_projector_stride from global config
* img_u8_resize_bilinear_to_f32 move in step3vl class
* Apply suggestions from code review
... | [
{
"path": "convert_hf_to_gguf.py",
"patch": "@@ -2219,10 +2219,10 @@ def set_gguf_parameters(self):\n self.image_size = self.find_vparam([\"image_size\"])\n self.gguf_writer.add_vision_image_size(self.image_size)\n self.gguf_writer.add_vision_patch_size(self.find_vparam([... | 2026-04-08T07:51:31 |
facebook/react | e41a8d6a1ea481be0fd1b4b3cbe4930919525546 | 2ae3592d29caf9aca2cff36e6089c1dfb92dfa4e | Validate destructuring assignment to globals
Fixes one more category of bug. For assignment expressions, we validating
against redeclaring a global variable when the assignment target was an
identifier, but not when the global was reassigned via destructuring. This PR
adds a `lowerIdentifierForAssignment()` helper... | [
{
"path": "compiler/forget/src/HIR/BuildHIR.ts",
"patch": "@@ -2439,6 +2439,41 @@ function getLoadKind(\n return isContext ? \"LoadContext\" : \"LoadLocal\";\n }\n \n+function lowerIdentifierForAssignment(\n+ builder: HIRBuilder,\n+ loc: SourceLocation,\n+ kind: InstructionKind,\n+ path: NodePath<t.Id... | 2023-06-04T18:29:55 |
huggingface/transformers | df1cd3a7bfe3e91a4c9bc32cd273d3683f976da6 | 147b7aa040812b079f467e777a2d2e1284167de0 | Models with incorrect tokenizer_class in tokenization_config.json tha… (#44179)
Models with incorrect tokenizer_class in tokenization_config.json that should use TokenziersBackend
Co-authored-by: ita.zaporozhets@huggingface.co <ita_zaporozhets@ip-26-0-164-75.ec2.internal> | [
{
"path": "src/transformers/models/auto/tokenization_auto.py",
"patch": "@@ -185,6 +185,12 @@\n (\"megatron-bert\", \"BertTokenizer\" if is_tokenizers_available() else None),\n (\"metaclip_2\", \"XLMRobertaTokenizer\" if is_tokenizers_available() else None),\n (\"mgp-str\", \"MgpstrT... | 2026-02-23T08:33:11 |
ggml-org/llama.cpp | 97508acb17ff933c67edb3a2f0997bc19a6abb98 | 5c4aae66e15990f87815c9eba1663f728067512b | webui: fix syntax highlighting lost after streaming for non-common languages (#21206)
* webui: fix syntax highlighting lost for non-common languages after streaming
rehype-highlight uses lowlight internally, which only bundles 37 "common"
languages. The streaming code path uses highlight.js directly (192 languages),
... | [
{
"path": "tools/server/public/index.html",
"patch": "@@ -18,7 +18,7 @@\n \t\t<div style=\"display: contents\">\n \t\t\t<script>\n \t\t\t\t{\n-\t\t\t\t\t__sveltekit_10avopp = {\n+\t\t\t\t\t__sveltekit_1ppa22i = {\n \t\t\t\t\t\tbase: new URL('.', location).pathname.slice(0, -1)\n \t\t\t\t\t};\n ",
"addit... | 2026-04-08T06:58:08 |
huggingface/transformers | a3bfc8bcd53bb8c86fb239dac55a8ba163b18f6a | ecf79eb2f61bda25c38a41ce6480edfdd569bec4 | fix(flaky): `test_generate_with_and_without_position_ids` in GLM ORC (#44173)
Fixes flaky GLM OCR generation behavior when 2D position_ids are passed explicitly. | [
{
"path": "tests/generation/test_utils.py",
"patch": "@@ -2260,17 +2260,31 @@ def test_forward_with_logits_to_keep(self):\n torch.testing.assert_close(all_logits[:, -1:, :], last_token_logits, rtol=1e-5, atol=1e-5)\n \n def test_generate_with_and_without_position_ids(self):\n+ ran_any... | 2026-02-20T19:06:17 |
electron/electron | 028a72daaccc6f9085ffd54a51c07bbf427c76fa | 6a1748da0608428fe902a3fc04a27899cd9803ad | build: fix macos release GN gen (#33915) | [
{
"path": "BUILD.gn",
"patch": "@@ -1119,7 +1119,7 @@ if (is_mac) {\n }\n \n extract_symbols(\"egl_syms\") {\n- binary = \"$root_out_dir/ibEGL.dylib\"\n+ binary = \"$root_out_dir/libEGL.dylib\"\n symbol_dir = \"$root_out_dir/breakpad_symbols\"\n dsym_file = \"$root_out_dir/libE... | 2022-04-25T21:00:32 |
golang/go | c4f356dd863c449835248c24f6dc653323010a1c | 08725f9de28726b52c2e62a37cd378d10f10c110 | crypto/ecdsa: fix s390x assembly with P-521
I had incorrectly assumed that the blocksize was always the same as the
curve field size. This is true of P-256 and P-384, but not P-521.
Fixes #70660
Fixes #70771
Change-Id: Idb6b510fcd3dd42d9b1e6cf42c1bb92e0ce8bd07
Reviewed-on: https://go-review.googlesource.com/c/go/+/6... | [
{
"path": "src/crypto/internal/fips140/ecdsa/ecdsa.go",
"patch": "@@ -21,7 +21,7 @@ import (\n \n type PrivateKey struct {\n \tpub PublicKey\n-\td []byte // bigmod.(*Nat).Bytes output (fixed length)\n+\td []byte // bigmod.(*Nat).Bytes output (same length as the curve order)\n }\n \n func (priv *PrivateK... | 2024-12-13T15:59:20 |
facebook/react | 2ae3592d29caf9aca2cff36e6089c1dfb92dfa4e | 6d6518161afdcbf7f2504441def52e3217abc869 | Fix/review bug repros, mark fixed
I reviewed the test cases we have marked as bugs ("_bug.*") and realized that
several of them are already fixed — woohoo! Then one wasn't fixed _yet_: our
type inference loses track of refs if you stash them inside an object/array. But
that's why I added the ValidateNoRefAccessInR... | [
{
"path": "compiler/forget/src/HIR/ValidateNoRefAccesInRender.ts",
"patch": "@@ -40,6 +40,15 @@ export function validateNoRefAccessInRender(fn: HIRFunction): void {\n for (const [, block] of fn.body.blocks) {\n for (const instr of block.instructions) {\n switch (instr.value.kind) {\n+ cas... | 2023-06-04T18:08:54 |
ggml-org/llama.cpp | 66c4f9ded01b29d9120255be1ed8d5835bcbb51d | 93bdc6156333082421d82b52d36f7dacb0542495 | ggml-cuda: ds_read_b128 for q4_0 and q4_1 mmq kernels (#21168)
* ds_read_b128 for q4_0 and q4_1 mmq kernels
Current for loop generates ds_read_b32 instructions with hip compiler, the new solution generates ds_read_b128 instructions for the same operation, saving some LDS bandwidth. Tested on MI50 and RX6800XT, i... | [
{
"path": "ggml/src/ggml-cuda/mmq.cuh",
"patch": "@@ -386,17 +386,25 @@ static __device__ __forceinline__ void vec_dot_q4_0_q8_1_dp4a(\n #pragma unroll\n for (int i0 = 0; i0 < mmq_y; i0 += warp_size) {\n const int i = i0 + threadIdx.x;\n-\n const int kyqs = QI8_1 ... | 2026-04-07T19:47:42 |
golang/go | 38e9a671d7648227f4f5b133e2e6452491cccebf | 6f7a4540b13d6d3be997276178aed96fb0e8a9c2 | syscall: on freebsd-386 only update written for certain errors
Testing on the freebsd-386 gomote seems to show that sendfile returns
a non-zero number of bytes written even when it returns EINVAL.
This confuses the caller. Change the Go code to only return non-zero
on success or EINTR or EAGAIN, which are the only cas... | [
{
"path": "src/syscall/syscall_freebsd_386.go",
"patch": "@@ -36,7 +36,13 @@ func sendfile(outfd int, infd int, offset *int64, count int) (written int, err e\n \tvar writtenOut uint64 = 0\n \t_, _, e1 := Syscall9(SYS_SENDFILE, uintptr(infd), uintptr(outfd), uintptr(*offset), uintptr((*offset)>>32), uintptr(... | 2024-12-12T22:07:13 |
huggingface/transformers | ecf79eb2f61bda25c38a41ce6480edfdd569bec4 | 708d3e125150569436f8db0ca608a9a47f876814 | [generate] Completely stop relying on `cache_position` to prepare inputs (#44130)
* generalize
* better doc
* don't mixup the logic everywhere
* doc
* fix all
* no unsqueeze
* more fixes
* add position_ids
* simplify all a lot
* doc
* small oupsi
* more small fixes
* more fixes
* fix assisted decoding
* ... | [
{
"path": "src/transformers/generation/candidate_generator.py",
"patch": "@@ -39,9 +39,7 @@\n class CandidateGenerator:\n \"\"\"Abstract base class for all candidate generators that can be applied during assisted generation.\"\"\"\n \n- def get_candidates(\n- self, input_ids: torch.LongTensor,... | 2026-02-20T18:46:17 |
ollama/ollama | c85c0ebf895016c36bab10be4dd92f594c400df3 | 10a8e04a8dcdd06de4ccaa3c2fe19452d2714b15 | CI: switch windows to vs 2022 (#11184)
* CI: switch windows to vs 2022
* ci: fix regex match | [
{
"path": ".github/workflows/release.yaml",
"patch": "@@ -155,6 +155,9 @@ jobs:\n echo \"$hipPath\\bin\" | Out-File -FilePath $env:GITHUB_PATH -Encoding utf8 -Append\n echo \"CC=$hipPath\\bin\\clang.exe\" | Out-File -FilePath $env:GITHUB_ENV -Append\n echo \"CXX=$hipPath\\bin\\... | 2025-06-24T20:26:55 |
facebook/react | f181ba8aa6339d62f6e2572109c61242606f16b3 | e1ad4aa3615333009d76f947ff05ddeff01039c6 | [Flight] Add bundler-less version of RSC using plain ESM (#26889)
This isn't really meant to be actually used, there are many issues with
this approach, but it shows the capabilities as a proof-of-concept.
It's a new reference implementation package `react-server-dom-esm` as
well as a fixture in `fixtures/flight-... | [
{
"path": ".eslintrc.js",
"patch": "@@ -324,6 +324,7 @@ module.exports = {\n 'packages/react-devtools-shared/**/*.js',\n 'packages/react-noop-renderer/**/*.js',\n 'packages/react-refresh/**/*.js',\n+ 'packages/react-server-dom-esm/**/*.js',\n 'packages/react-server-dom... | 2023-06-03T19:58:24 |
ggml-org/llama.cpp | 93bdc6156333082421d82b52d36f7dacb0542495 | 4eb19514dd2984662f13aacbb052c559c8fde3b1 | gguf-py : fix missing comma after bad merge in tensor-mapping (#21558)
This commit adds a missing comma in the vision encoder attention qkv
block.
The motivation for this change is that without the comma there will be
a string concatenation of the Kimi-K2.5 and the Nemotron Nano v2 VL
tensor mappings which will be br... | [
{
"path": "gguf-py/gguf/tensor_mapping.py",
"patch": "@@ -1441,7 +1441,7 @@ class TensorNameMap:\n \"visual.blocks.{bid}.attn.qkv\", # qwen3vl\n \"model.vision.transformer.layers.{bid}.attention.query_key_value\", # cogvlm\n \"model.vision_model.transformer.layers.{bid}.s... | 2026-04-07T19:24:25 |
golang/go | 6f7a4540b13d6d3be997276178aed96fb0e8a9c2 | 14e5093ee56c7e3a807c8924fd2d425cd2b0f376 | net: fix example function name for IP.To4
Change-Id: Ia9a2c3a9f53792173cd1fb9f8e1a078fe3444945
Reviewed-on: https://go-review.googlesource.com/c/go/+/635136
Auto-Submit: Ian Lance Taylor <iant@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Carlos Amed... | [
{
"path": "src/net/example_test.go",
"patch": "@@ -334,7 +334,7 @@ func ExampleIP_To16() {\n \t// 10.255.0.0\n }\n \n-func ExampleIP_to4() {\n+func ExampleIP_To4() {\n \tipv6 := net.IP{0xfc, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0}\n \tipv4 := net.IPv4(10, 255, 0, 0)\n ",
"additions": 1,
"deleti... | 2024-12-12T18:08:35 |
huggingface/transformers | 87954470edab2b03b56bbb6bf530a2f07f9b2180 | 7c0a3be0bbe0d19bdbd1e59259ee06684f05fa58 | [`Mamba`] Fix kernel loading (#44176)
* fix
* fix all imports
* falcon h1
* fix check conditions and fixup zamba | [
{
"path": "src/transformers/models/bamba/modeling_bamba.py",
"patch": "@@ -44,6 +44,7 @@\n from ...processing_utils import Unpack\n from ...utils import TransformersKwargs, auto_docstring, can_return_tuple, is_torchdynamo_compiling, logging\n from ...utils.generic import maybe_autocast\n+from ...utils.impor... | 2026-02-20T16:18:58 |
ollama/ollama | 1c6669e64cc8a482fbf1e35c0249f17b35a4e87a | 2bb69b40c7f6f2783290f0b7e4f7d5ec0a41f69c | Re-remove cuda v11 (#10694)
* Re-remove cuda v11
Revert the revert - drop v11 support requiring drivers newer than Feb 23
This reverts commit c6bcdc4223c50071b59a19c42cc54ec9932f696f.
* Simplify layout
With only one version of the GPU libraries, we can simplify things down somewhat. (Jetsons still require special... | [
{
"path": ".github/workflows/release.yaml",
"patch": "@@ -103,11 +103,6 @@ jobs:\n arch: [amd64]\n preset: ['CPU']\n include:\n- - os: windows\n- arch: amd64\n- preset: 'CUDA 11'\n- install: https://developer.download.nvidia.com/compute/cuda/... | 2025-06-23T21:07:00 |
facebook/react | e1ad4aa3615333009d76f947ff05ddeff01039c6 | 5fb2c15a89de844a1dd12a61e7674e55dc0dfa89 | [Fizz][Float] stop automatically preloading scripts that are not script resources (#26877)
Currently we preload all scripts that are not hoisted. One of the
original reasons for this is we stopped SSR rendering async scripts that
had an onLoad/onError because we needed to be able to distinguish
between Float script... | [
{
"path": "packages/react-dom-bindings/src/client/ReactFiberConfigDOM.js",
"patch": "@@ -1036,19 +1036,6 @@ export function bindInstance(\n \n export const supportsHydration = true;\n \n-// With Resources, some HostComponent types will never be server rendered and need to be\n-// inserted without breaking h... | 2023-06-01T20:34:36 |
ggml-org/llama.cpp | 957d717ce54401a28591c265d19c53ba620f4c46 | de1aa6fa73e135839109e09fec1a0997f4207b2a | ggml-webgpu: parameterize submission size and add iOS specific limits (#21533)
* Work towards removing bitcast
* Move rest of existing types over
* Add timeout back to wait and remove synchronous set_tensor/memset_tensor
* move to unpackf16 for wider compatibility
* cleanup
* Remove deadlock condition in free_buf... | [
{
"path": "ggml/src/ggml-webgpu/ggml-webgpu.cpp",
"patch": "@@ -16,7 +16,6 @@\n #include <webgpu/webgpu_cpp.h>\n \n #include <atomic>\n-#include <condition_variable>\n #include <cstdint>\n #include <cstring>\n #ifdef GGML_WEBGPU_GPU_PROFILE\n@@ -25,7 +24,6 @@\n #if defined(GGML_WEBGPU_DEBUG) || defined(GGML... | 2026-04-07T17:30:01 |
golang/go | 077d51909d3d7bc2d52afd47c9be1de8ee4f0756 | fafd4477f3d19f2c11a628e6e407ecf9924309c1 | internal/poll: in SendFile treat ENOTSUP like EOPNOTSUPP
Fixes #70763
Change-Id: Ifb79b5b0529f7977df0fe1b59d224b8b31df2c9b
Reviewed-on: https://go-review.googlesource.com/c/go/+/635396
Reviewed-by: Ian Lance Taylor <iant@google.com>
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.... | [
{
"path": "src/internal/poll/sendfile_unix.go",
"patch": "@@ -110,12 +110,20 @@ func sendFile(dstFD *FD, src int, offset *int64, size int64) (written int64, err\n \t\t\t// Retry.\n \t\tcase syscall.ENOSYS, syscall.EOPNOTSUPP, syscall.EINVAL:\n \t\t\t// ENOSYS indicates no kernel support for sendfile.\n-\t\t... | 2024-12-11T22:02:28 |
huggingface/transformers | 9887ca2ab5fc922e5f5c85070b5028c0284fbb44 | d23397091ccc5aa0e647b5a850e1bbde1989fd2d | ESM2 attention_mask and token_dropout fix (#44163)
* Update modeling_esm.py
* Update modeling_esm.py
* Skip irrelevant test
---------
Co-authored-by: Matt <rocketknight1@gmail.com> | [
{
"path": "src/transformers/models/esm/modeling_esm.py",
"patch": "@@ -644,8 +644,11 @@ def forward(\n raise ValueError(\"You must specify exactly one of input_ids or inputs_embeds\")\n \n if inputs_embeds is None:\n+ # Important, attention_mask must be passed to the embedding... | 2026-02-20T15:16:57 |
facebook/react | 811022232efed62a3c943701bc99d18655bc78b3 | 018c58c9c65452cff25aaf1f38f78a9b90d8e5c1 | Always trigger componentWillUnmount in StrictMode (#26842)
<!--
Thanks for submitting a pull request!
We appreciate you spending the time to work on these changes. Please
provide enough information so that others can review your pull request.
The three fields below are mandatory.
Before submitting a pull requ... | [
{
"path": "packages/react-reconciler/src/ReactFiberClassComponent.js",
"patch": "@@ -10,7 +10,6 @@\n import type {Fiber} from './ReactInternalTypes';\n import type {Lanes} from './ReactFiberLane';\n import type {UpdateQueue} from './ReactFiberClassUpdateQueue';\n-import type {Flags} from './ReactFiberFlags'... | 2023-06-01T17:34:31 |
ggml-org/llama.cpp | 69c28f1547c169902f62ca48bee75fb876c4d8e6 | 0d049d6a9245ccffc6073743ff8b6bb24ac6a47b | llama-server: fix model params not propagated (#21509)
Signed-off-by: Aaron Teo <aaron.teo1@ibm.com> | [
{
"path": "tools/server/server-context.cpp",
"patch": "@@ -632,7 +632,7 @@ struct server_context_impl {\n \n // load the model and initialize llama_context\n // this may also be called to resume from sleeping state\n- bool load_model(const common_params & params) {\n+ bool load_model(common_pa... | 2026-04-07T13:39:41 |
ollama/ollama | 0a066cfd91abdddc6ee172776974a6720a3072d3 | 87b7af6ceef2b4d96374dbff5070b41b17d3f138 | Reapply "feat: incremental gguf parser (#10822)" (#11114) (#11119)
* Reapply "feat: incremental gguf parser (#10822)" (#11114)
This reverts commit a6e64fbdf28f0d6cb97cc7f022ca493b905fe895.
* fix older ggufs | [
{
"path": "fs/gguf/gguf.go",
"patch": "@@ -0,0 +1,347 @@\n+package gguf\n+\n+import (\n+\t\"bytes\"\n+\t\"cmp\"\n+\t\"encoding/binary\"\n+\t\"errors\"\n+\t\"fmt\"\n+\t\"io\"\n+\t\"iter\"\n+\t\"os\"\n+\t\"slices\"\n+\t\"strings\"\n+)\n+\n+const (\n+\ttypeUint8 uint32 = iota\n+\ttypeInt8\n+\ttypeUint16\n+\tty... | 2025-06-20T18:11:40 |
golang/go | fafd4477f3d19f2c11a628e6e407ecf9924309c1 | 0ca521f9c1a6357578b501615f073adb2da6b527 | cmd/cgo: use full prototype for main in C code
Avoids pedantic errors from modern C compilers.
Fixes #70769
Change-Id: Ibe0d9960e998eb0346b97d7bd69eb7de0d0e6c60
Reviewed-on: https://go-review.googlesource.com/c/go/+/635095
Reviewed-by: Robert Griesemer <gri@google.com>
Reviewed-by: Mauri de Souza Meneguzzo <mauri870... | [
{
"path": "src/cmd/cgo/doc.go",
"patch": "@@ -796,7 +796,7 @@ Instead, the build process generates an object file using dynamic\n linkage to the desired libraries. The main function is provided by\n _cgo_main.c:\n \n-\tint main() { return 0; }\n+\tint main(int argc, char **argv) { return 0; }\n \tvoid cross... | 2024-12-11T04:05:06 |
huggingface/transformers | 32deb4c32a9b5c9f455199f5a076f1112fc89f87 | 8151000f745deb056973e3b1387b4a537a586d63 | Fix `Seq2SeqTrainingArguments` documentation (#35258)
* try fix Seq2SeqTrainingArguments
* ignore formatter | [
{
"path": "src/transformers/training_args_seq2seq.py",
"patch": "@@ -28,7 +28,12 @@\n @add_start_docstrings(TrainingArguments.__doc__)\n class Seq2SeqTrainingArguments(TrainingArguments):\n \"\"\"\n- Args:\n+ sortish_sampler (`bool`, *optional*, defaults to `False`):\n+ Whether to u... | 2026-02-20T13:49:31 |
electron/electron | f3e0517b6e1ddb0fbd05949007720d685fc86924 | 53c3dd68b29e2a864aff9cb6d758a727e5665972 | chore: bump chromium to 102.0.4999.0 (main) (#33731)
* chore: bump chromium in DEPS to 102.0.4999.0
* 3576640: Set OOM handler during V8 initialization
https://chromium-review.googlesource.com/c/chromium/src/+/3576640
* 3574964: Remove deprecated base::Value usage in print_settings_conversion code.
https:/... | [
{
"path": ".circleci/config/base.yml",
"patch": "@@ -24,7 +24,7 @@ parameters:\n linux-publish-arch-limit:\n type: enum\n default: all\n- enum: [\"all\", \"arm\", \"arm64\", \"x64\", \"ia32\"]\n+ enum: [\"all\", \"arm\", \"arm64\", \"x64\"]\n \n run-macos-publish:\n type: boolean\n@@ -... | 2022-04-22T22:36:22 |
ggml-org/llama.cpp | 0d049d6a9245ccffc6073743ff8b6bb24ac6a47b | a8ec0df4617692ef0f18b212f2e16bba540e053a | unicode : add custom Qwen2 regex handler to fix segfault on long input (#21257)
* unicode : add custom Qwen2 regex handler to fix segfault on long input
std::regex uses recursive backtracking internally, which causes a stack
overflow (segfault) when tokenizing long sequences of repeated characters
(e.g. 43K 'A's). Th... | [
{
"path": "src/unicode.cpp",
"patch": "@@ -470,6 +470,141 @@ static std::vector<size_t> unicode_regex_split_custom_llama3(const std::string &\n return bpe_offsets;\n }\n \n+// Qwen2 system regex: \"(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\\\r\\\\n\\\\p{L}\\\\p{N}]?\\\\p{L}+|\\\\p{N}| ?[^\\\\s\\\\p{L}\\\\p{N}]+[... | 2026-04-07T13:13:38 |
facebook/react | 2c8f6888e68b0f8d0d83c6563efbdd3dd25c8feb | 91f41f5ee321fe4cebb5f33bf524bbe52fe9b24b | [hir] Failing test for adding ref as dep incorrectly | [
{
"path": "compiler/forget/src/__tests__/fixtures/compiler/_bug.use-ref-added-to-dep-without-type-info.expect.md",
"patch": "@@ -0,0 +1,61 @@\n+\n+## Input\n+\n+```javascript\n+function Foo({ a }) {\n+ const ref = useRef();\n+ // type information is lost here as we don't track types of fields\n+ const va... | 2023-05-31T12:55:00 |
ollama/ollama | 87b7af6ceef2b4d96374dbff5070b41b17d3f138 | f2527b08fba57d606e12cb21b583249c11724d7a | ggml: Check return status for computation.
We don't check the return status after computing the graph, which
can silently lead to bad outputs if we try to keep going and future
computation succeeds. This appears to happens in certain cases on
Apple M2 devices.
Fixes #11070 | [
{
"path": "ml/backend/ggml/ggml.go",
"patch": "@@ -602,7 +602,9 @@ func (c *Context) Forward(tensors ...ml.Tensor) ml.Context {\n }\n \n func (c *Context) Compute(tensors ...ml.Tensor) {\n-\tC.ggml_backend_sched_graph_compute_async(c.b.sched, c.graph)\n+\tif status := C.ggml_backend_sched_graph_compute_asyn... | 2025-06-19T21:39:20 |
huggingface/transformers | e0b07d1b120fdc374b9a563f7edf0305401216be | 64230197ff9a4499aa8557eb5b225d164ad8d7c7 | [`simple`] Fix up `__repr__` whitespace/brackets (#44048)
Fix up __repr__ whitespace/brackets | [
{
"path": "src/transformers/tokenization_mistral_common.py",
"patch": "@@ -1595,7 +1595,7 @@ def __repr__(self) -> str:\n f\"{self.__class__.__name__}(name_or_path='{self.name_or_path}',\"\n f\" vocab_size={self.vocab_size}, model_max_length={self.model_max_length},\"\n f... | 2026-02-20T10:03:33 |
golang/go | 0ca521f9c1a6357578b501615f073adb2da6b527 | c93477b5e563dd0ed7b45fd519762f24b7cfa7b0 | debug/elf: adjust version API per issue discussion
This updates the new version API for the discussion on #63952.
This change reveals that in fact none of the tests set the
VERSYM_HIDDEN bit. The code before this CL set the hidden flag
for symbols that appear in DynamicVersionNeed, but that is not
an accurate represe... | [
{
"path": "api/go1.24.txt",
"patch": "@@ -106,33 +106,34 @@ pkg debug/elf, const VER_FLG_INFO = 4 #63952\n pkg debug/elf, const VER_FLG_INFO DynamicVersionFlag #63952\n pkg debug/elf, const VER_FLG_WEAK = 2 #63952\n pkg debug/elf, const VER_FLG_WEAK DynamicVersionFlag #63952\n-pkg debug/elf, const VerFlagGl... | 2024-12-11T03:48:11 |
electron/electron | 33e9bfd99bd6b8802aa60fefcf3c87df2341b837 | e864bc44c66de68e6b1645a41e79a1eda77538c2 | build: improve circleci config (#33881)
* build: fix conditional restore of git cache
* build: split lint out of setup.yml | [
{
"path": ".circleci/.gitignore",
"patch": "@@ -0,0 +1 @@\n+config-staging",
"additions": 1,
"deletions": 0,
"language": "Unknown"
},
{
"path": ".circleci/config.yml",
"patch": "@@ -6,6 +6,7 @@ setup: true\n # Orbs\n orbs:\n path-filtering: circleci/path-filtering@0.1.0\n+ continu... | 2022-04-21T21:34:25 |
ggml-org/llama.cpp | e8f50826979b5d3d4f54e522e552caf430edce0d | 22fc79134e23feea5cbc1074eaae77143932ccbc | server : fix restore for checkpoints with pos_min == 0 (#21510) | [
{
"path": "tools/server/server-context.cpp",
"patch": "@@ -2404,7 +2404,7 @@ struct server_context_impl {\n // guarantee that a checkpoint will result in at least one token being processed [TAG_PROMPT_LOGITS]\n LOG_INF(\... | 2026-04-07T12:29:17 |
facebook/react | fc13557d30c50517f5f1d369ce2ee02f506bb904 | 71c9743e388999a5ffe5997fe259a2af3283dad3 | [tests] remove jest fixture tests
---
Remove jest fixture tests in favor of snap runner. Main reasons:
- maintaining feature flags and compatible behavior required syncing all changes
to 3 files (`generateTestsFromFixtures`, `compiler-test`, and `compiler-worker`)
- jest snapshot test file causes rebase confli... | [
{
"path": "compiler/forget/src/__tests__/compiler-test.ts",
"patch": "@@ -1,138 +0,0 @@\n-/**\n- * Copyright (c) Meta Platforms, Inc. and affiliates.\n- *\n- * This source code is licensed under the MIT license found in the\n- * LICENSE file in the root directory of this source tree.\n- */\n-\n-\"use strict... | 2023-05-26T17:18:38 |
huggingface/transformers | 64230197ff9a4499aa8557eb5b225d164ad8d7c7 | 1f468a8b5d55397e72135d90cf5e9e3b710ebc2a | [`chore`] Fix incorrect forward type hint for Gemma3n (#44051)
* Fix incorrect forward type hint for Gemma3n
* Also on modular | [
{
"path": "src/transformers/models/gemma3n/modeling_gemma3n.py",
"patch": "@@ -2034,7 +2034,7 @@ def forward(\n output_attentions: bool | None = None,\n output_hidden_states: bool | None = None,\n **lm_kwargs: Unpack[TransformersKwargs],\n- ) -> Gemma3nCausalLMOutputWithPast:\n+ ... | 2026-02-20T09:05:33 |
ollama/ollama | 60cfa2a203201e8513b7da26035f74495e498bc7 | 55bbf3b4a1766fc99e32ebd39f0dd5749152b55d | cache: fix comment function name in cache.go (#11110) | [
{
"path": "server/internal/cache/blob/cache.go",
"patch": "@@ -59,7 +59,7 @@ type DiskCache struct {\n \ttestHookBeforeFinalWrite func(f *os.File)\n }\n \n-// PutString is a convenience function for c.Put(d, strings.NewReader(s), int64(len(s))).\n+// PutBytes is a convenience function for c.Put(d, strings.N... | 2025-06-18T12:21:45 |
electron/electron | 64517b36efb06464fcf6d2f214fd75d833fdb07a | 00021a41b1905cbee93e775958c4fb50d473cda4 | build: fix intermittent compilation failures on macOS (#33768)
* build: fix intermittent compilation failures on macOS
* chore: remove //base dependency from main executable
* chore: fix lint
Co-authored-by: deepak1556 <hop2deep@gmail.com> | [
{
"path": "BUILD.gn",
"patch": "@@ -1072,7 +1072,6 @@ if (is_mac) {\n \"shell/app/electron_main_mac.cc\",\n \"shell/app/uv_stdio_fix.cc\",\n \"shell/app/uv_stdio_fix.h\",\n- \"shell/common/electron_constants.cc\",\n ]\n include_dirs = [ \".\" ]\n deps = [",
"additions"... | 2022-04-21T08:25:51 |
golang/go | c93477b5e563dd0ed7b45fd519762f24b7cfa7b0 | 3104b6adbb36a43284f51ab0cb67c44f8ba75fac | crypto: use provided random Reader in FIPS mode
This removes the difference in behavior between FIPS mode on and off.
Instead of the sentinel type we could have moved the Reader to the
drbg package and checked for equality, but then we would have locked the
crypto/rand.Reader implementation to the one in the FIPS mod... | [
{
"path": "src/crypto/ecdh/nist.go",
"patch": "@@ -8,6 +8,7 @@ import (\n \t\"bytes\"\n \t\"crypto/internal/boring\"\n \t\"crypto/internal/fips140/ecdh\"\n+\t\"crypto/internal/fips140only\"\n \t\"errors\"\n \t\"io\"\n )\n@@ -43,6 +44,10 @@ func (c *nistCurve) GenerateKey(rand io.Reader) (*PrivateKey, error)... | 2024-12-11T13:50:00 |
ggml-org/llama.cpp | 2a619f6fbce255924e8c08a2b82ff96e1d65c2e6 | edd4d9bca5dcb9be1ce66e618157d8787a0ea16f | ggml: Vulkan build, Linux -- output error string for errno on fork failure (#20868) (#20904) | [
{
"path": "ggml/src/ggml-vulkan/vulkan-shaders/vulkan-shaders-gen.cpp",
"patch": "@@ -137,6 +137,7 @@ void execute_command(std::vector<std::string>& command, std::string& stdout_str,\n \n pid_t pid = fork();\n if (pid < 0) {\n+ std::cerr << strerror(errno) << \"\\n\";\n throw std::run... | 2026-04-07T11:54:55 |
facebook/react | 71c9743e388999a5ffe5997fe259a2af3283dad3 | 99257aa4303c09186ca9a99fe0cf4a5995d4bd96 | [snap] Fix vscode debugger attaching to forked proc
Turns out I just forgot to forward `process.env` when forking 😅
`debugger` statements and breakpoints should work with both `--sync` and
`--no-sync` (default) modes | [
{
"path": "compiler/forget/packages/snap/src/main.ts",
"patch": "@@ -34,7 +34,9 @@ const childProc = fork(require.resolve(\"./runner.js\"), hideBin(process.argv), {\n // when we `inherit` stdin.\n // pipe stdout and stderr so we can silence child process after parent exits\n stdio: [\"pipe\", \"pipe\"... | 2023-05-26T17:18:37 |
ollama/ollama | 6bda1d24798e40fc9ea1419c6ce22c6cdcc9dfe2 | 9e125d884cf995dfae7fcd74690d525e4326a517 | tools: fix parsing tool calls without any parameters (#11101)
Fixes issue where tool calls that don't expect any parameters were
not being parsed. This also fixes two additional issues: one where
2+ tool calls would not be correctly parsed, and cases where tool calls
with invalid parameters would still get parsed | [
{
"path": "tools/tools.go",
"patch": "@@ -18,9 +18,8 @@ const (\n )\n \n type Parser struct {\n-\ttag string\n-\tnames []string\n-\tproperties []string\n+\ttag string\n+\ttools []api.Tool\n \n \tstate toolsState\n \tbuffer []byte\n@@ -34,15 +33,10 @@ func NewParser(tmpl *template.Template, to... | 2025-06-17T17:51:43 |
huggingface/transformers | 1f468a8b5d55397e72135d90cf5e9e3b710ebc2a | 6b55285d93f7cd6227d90311c314cbab05b2cbfd | Raise informative error when loading video processors (#44125)
* fix
* move it to the last spot | [
{
"path": "src/transformers/models/auto/video_processing_auto.py",
"patch": "@@ -94,8 +94,8 @@\n \n \n def video_processor_class_from_name(class_name: str):\n- for module_name, extractors in VIDEO_PROCESSOR_MAPPING_NAMES.items():\n- if class_name in extractors:\n+ for module_name, extractor in ... | 2026-02-20T08:23:35 |
ggml-org/llama.cpp | 71a81f6fcc2c7e4bf17c3c2484c9498358d173b2 | ecce0087da749a5d7b0314aae9f1f136eee8b834 | ggml-cuda : fix CDNA2 compute capability constant for gfx90a (MI210) (#21519)
GGML_CUDA_CC_CDNA2 was set to 0x910
Fix by setting the constant to 0x90a to match the actual gfx90a ISA. | [
{
"path": "ggml/src/ggml-cuda/common.cuh",
"patch": "@@ -65,7 +65,7 @@\n #define GGML_CUDA_CC_VEGA (GGML_CUDA_CC_OFFSET_AMD + 0x900) // Vega56/64, minimum for fp16 dual issue\n #define GGML_CUDA_CC_VEGA20 (GGML_CUDA_CC_OFFSET_AMD + 0x906) // MI50/Radeon VII, minimum for dp4a\n #define GGML_CUDA_... | 2026-04-07T10:18:55 |
electron/electron | b53118ca28bcc9dcff3304f33d7c576a82b3f298 | 0c864837af77d6d8a59b6c44f982dcafc3d2fdde | fix: adopt partition alloc early initialization (#33832)
Refs https://chromium-review.googlesource.com/c/chromium/src/+/3298858 | [
{
"path": "BUILD.gn",
"patch": "@@ -916,7 +916,10 @@ if (is_mac) {\n assert(defined(invoker.helper_name_suffix))\n \n output_name = electron_helper_name + invoker.helper_name_suffix\n- deps = [ \":electron_framework+link\" ]\n+ deps = [\n+ \":electron_framework+link\",\n+ ... | 2022-04-20T20:52:15 |
facebook/react | 4daccade04ae54339463546e06f07db56a646bd4 | 18dedde6a7d63ffafd2f7e135f9edf424eee0bcd | Remove temporary CircleCI workaround (#26855)
There was a CircleCI bug that prevented the sizebot job from accessing
the artifacts API. I had added a temporary workaround to pull from a
mirror instead. This seems to have been fixed, so I can remove the
workaround. | [
{
"path": ".circleci/config.yml",
"patch": "@@ -151,32 +151,11 @@ jobs:\n - setup_node_modules\n - run:\n name: Download artifacts for base revision\n- # TODO: The download-experimental-build.js script works by fetching\n- # artifacts from CI. CircleCI recently update... | 2023-05-25T17:38:39 |
golang/go | 3104b6adbb36a43284f51ab0cb67c44f8ba75fac | 5424f2e200e022e5ddf95088118fb0914343492a | log/slog: make DiscardHandler example package-level
Fixes #70782.
Change-Id: I8e8b763040bd10147eb7d1a30ac0774e28f90911
Reviewed-on: https://go-review.googlesource.com/c/go/+/635217
LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com>
Reviewed-by: Ian Lance Taylor <iant@google.com> | [
{
"path": "src/log/slog/example_discard_test.go",
"patch": "@@ -10,7 +10,7 @@ import (\n \t\"os\"\n )\n \n-func ExampleDiscardHandler() {\n+func Example_discardHandler() {\n \t// A slog.TextHandler can output log messages.\n \tlogger1 := slog.New(slog.NewTextHandler(\n \t\tos.Stdout,",
"additions": 1,
... | 2024-12-11T19:38:57 |
ollama/ollama | a6fbfc880c3de9b57e341db374907e2fedda9fa6 | 502028968ddca04bd19c0859a73fb4e0cbeac3e1 | gguf: fix write order (#11068)
* ggml: test write gguf order
* ggml: fix write tensor order | [
{
"path": "fs/ggml/gguf.go",
"patch": "@@ -527,23 +527,17 @@ func WriteGGUF(f *os.File, kv KV, ts []*Tensor) error {\n \t\treturn err\n \t}\n \n-\tkeys := slices.Collect(maps.Keys(kv))\n-\tslices.Sort(keys)\n-\n-\tfor _, key := range keys {\n+\tfor _, key := range slices.Sorted(maps.Keys(kv)) {\n \t\tif err... | 2025-06-16T17:42:32 |
huggingface/transformers | 6b55285d93f7cd6227d90311c314cbab05b2cbfd | 1618d44b9295361607ec74d7be860ba886aac039 | fix(flaky): Different approach to make sure loss exists (#43804)
The test now uses a deterministic approach and that fixes the reproducible (40%) flakiness | [
{
"path": "tests/models/ernie4_5_moe/test_modeling_ernie4_5_moe.py",
"patch": "@@ -19,6 +19,7 @@\n import pytest\n \n from transformers import BitsAndBytesConfig, is_torch_available\n+from transformers.models.ernie4_5_moe.modeling_ernie4_5_moe import load_balancing_loss_func\n from transformers.testing_util... | 2026-02-20T07:45:25 |
ggml-org/llama.cpp | ecce0087da749a5d7b0314aae9f1f136eee8b834 | d1f82e382ddf67f63ee0fd28ae613456cbb23481 | fix: Detect streaming state in reasoning content blocks (#21549) | [
{
"path": "tools/server/public/index.html",
"patch": "@@ -18,7 +18,7 @@\n \t\t<div style=\"display: contents\">\n \t\t\t<script>\n \t\t\t\t{\n-\t\t\t\t\t__sveltekit_1y361v9 = {\n+\t\t\t\t\t__sveltekit_1wqaxod = {\n \t\t\t\t\t\tbase: new URL('.', location).pathname.slice(0, -1)\n \t\t\t\t\t};\n ",
"addit... | 2026-04-07T10:04:41 |
facebook/react | 9a72e622716f6dab714ebe8b957ee9970154664b | ee4233bdbc71a7e09395a613c7dde01194d2a830 | run SchedulerFeatureFlags with variant flags again (#26851)
With df12d7eac40c87bd5fdde0aa5a739bce9e7dce27 I accidentally made it so
that tests aren't run with the 2 variant modes for most
SchedulerFeatureFlags anymore. This fixes it with the same approach as
ee4233bdbc71a7e09395a613c7dde01194d2a830.
Test Plan:
... | [
{
"path": "scripts/jest/setupTests.www.js",
"patch": "@@ -24,9 +24,18 @@ jest.mock('scheduler/src/SchedulerFeatureFlags', () => {\n ),\n {virtual: true}\n );\n- return jest.requireActual(\n+ const actual = jest.requireActual(\n schedulerSrcPath + '/src/forks/SchedulerFeatureFlags.www'\n ... | 2023-05-25T14:38:33 |
golang/go | a7c4cadce0799a74e48d394ff662ed5128667621 | 979c1cfbe8880e302d5a73df47f4efc3d34ee416 | cmd/compile: update broken link
Fixes #70778
Change-Id: Ie5ed53aa39446beb0316eb134cc705ea06b37435
Reviewed-on: https://go-review.googlesource.com/c/go/+/635295
Auto-Submit: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Keith Randall <khr@google.com>
Reviewed-by: Ian Lance... | [
{
"path": "src/cmd/compile/internal/ssa/_gen/rulegen.go",
"patch": "@@ -5,7 +5,8 @@\n // This program generates Go code that applies rewrite rules to a Value.\n // The generated code implements a function of type func (v *Value) bool\n // which reports whether if did something.\n-// Ideas stolen from Swift:... | 2024-12-11T16:59:04 |
electron/electron | 841e0a4e0ccae2b9fdac3944f7db1f23cb6a3c0a | 31c2b5703a75a5ff2bf0e68f472d9002fe866bfb | fix: include accessibility blink strings (#33840) | [
{
"path": "electron_paks.gni",
"patch": "@@ -184,6 +184,7 @@ template(\"electron_paks\") {\n \"${root_gen_dir}/components/strings/components_strings_\",\n \"${root_gen_dir}/device/bluetooth/strings/bluetooth_strings_\",\n \"${root_gen_dir}/services/strings/services_strings_\",\n+ \"${... | 2022-04-20T03:00:51 |
ollama/ollama | 82ad1dbc07dc2db39c6f502eb148ff3ce00d96b8 | feeabdadd2b272b40747f3e7e74957c40ba2800c | mac: handle "keep" named apps (#11031)
When a user elects to keep the existing app, the
new Ollama is named `Ollama 2.app`
This fixes the app startup flow to handle this naming pattern. | [
{
"path": "cmd/start_darwin.go",
"patch": "@@ -5,7 +5,7 @@ import (\n \t\"errors\"\n \t\"os\"\n \t\"os/exec\"\n-\t\"strings\"\n+\t\"regexp\"\n \n \t\"github.com/ollama/ollama/api\"\n )\n@@ -19,11 +19,12 @@ func startApp(ctx context.Context, client *api.Client) error {\n \tif err != nil {\n \t\treturn err\n ... | 2025-06-09T23:29:57 |
golang/go | 979c1cfbe8880e302d5a73df47f4efc3d34ee416 | e424d78c3da1732b72f9170e7c01f400926143ce | net: avoid unnecessary interface lookup fetching all interface addresses
InterfaceAddrs returns a list of the system's unicast interface addresses.
In order to do so, the function reuses the existing helpers and
list first all addresses with the netlink call RTM_GETADDR, then
all interfaces with RTM_GETLINK, and late... | [
{
"path": "src/net/interface_linux.go",
"patch": "@@ -129,22 +129,14 @@ func interfaceAddrTable(ifi *Interface) ([]Addr, error) {\n \tif err != nil {\n \t\treturn nil, os.NewSyscallError(\"parsenetlinkmessage\", err)\n \t}\n-\tvar ift []Interface\n-\tif ifi == nil {\n-\t\tvar err error\n-\t\tift, err = inte... | 2024-12-11T15:03:59 |
Subsets and Splits
Assembly Language GitHub Issues
Retrieves a sample of assembly-language related commits with their details, but doesn't provide meaningful analysis or patterns beyond basic filtering.
Swift Compiler Issues Analysis
Retrieves all training data for the Swift programming language repository, providing basic filtering but offering limited analytical insight beyond identifying relevant code examples.