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 |
|---|---|---|---|---|---|
vuejs/vue | 104b9406899299e3e18284b6fe4707dc93c9813e | fe02bc3a27b3b33f9467980f61126edf97105287 | handle component root patch edge case (transition + keep-alive + HOC) (fix #4590) | [
{
"path": "src/core/vdom/patch.js",
"patch": "@@ -178,6 +178,23 @@ export function createPatchFunction (backend) {\n }\n }\n \n+ function initComponent (vnode, insertedVnodeQueue) {\n+ if (vnode.data.pendingInsert) {\n+ insertedVnodeQueue.push.apply(insertedVnodeQueue, vnode.data.pendingInser... | 2017-01-12T23:25:20 |
huggingface/transformers | b3e3c3dc93f29770a768d6943c9fb9d377e5edce | b84c0b31c674436b076bba818885226a0e0ddecc | [Qwen3VL] fix device mismatch error for FSDP2 training (#41536)
For FSDP2, parameters might be on a meta device, and the weight.device attribute may
not accurately reflect where the actual computation will happen during forward passes.
```log
File "transformers/models/qwen3_vl_moe/modeling_qwen3_vl_moe.py", line 77... | [
{
"path": "src/transformers/models/qwen3_omni_moe/modeling_qwen3_omni_moe.py",
"patch": "@@ -1099,6 +1099,7 @@ def rot_pos_emb(self, grid_thw: torch.Tensor) -> torch.Tensor:\n \n def fast_pos_embed_interpolate(self, grid_thw):\n grid_ts, grid_hs, grid_ws = grid_thw[:, 0], grid_thw[:, 1], grid_th... | 2025-10-14T10:28:25 |
ggml-org/llama.cpp | 638d330246954e88dffc22ce01fec15e6894e544 | 84c8e305e8010a1a3d43bdd0a25f737ac67809a4 | ggml : fix graph reallocation with multiple chunks (#16396)
reallocation is needed if a single chunk grows in size,
even if total allocation size stays the same or is lower | [
{
"path": "ggml/src/ggml-alloc.c",
"patch": "@@ -392,12 +392,8 @@ static void ggml_dyn_tallocr_free(struct ggml_dyn_tallocr * alloc) {\n free(alloc);\n }\n \n-static size_t ggml_dyn_tallocr_max_size(struct ggml_dyn_tallocr * alloc) {\n- size_t max_size = 0;\n- for (int i = 0; i < alloc->n_chunks; ... | 2025-10-03T11:49:08 |
denoland/deno | df289e57f7f3d03513cb781dc2ee23c2b44d8c19 | 76da1f08dfcb165400a470b39a36d9f66f0825bb | fix(ext/node): properly implement process.features (#31864)
This almost correctly implements `process.features`, the only missing
bit is `process.features.inspector` which will be handled in
https://github.com/denoland/deno/pull/31863 because it requires to
disable a bunch of tests that were false positives.
Makes `p... | [
{
"path": "ext/node/polyfills/process.ts",
"patch": "@@ -94,7 +94,7 @@ import * as uv from \"ext:deno_node/internal_binding/uv.ts\";\n import type { BindingName } from \"ext:deno_node/internal_binding/mod.ts\";\n import { buildAllowedFlags } from \"ext:deno_node/internal/process/per_thread.mjs\";\n \n-const... | 2026-01-16T16:15:45 |
vuejs/vue | b708c04988e99f17e821dc82fa33faaeb92dc4a6 | 7131bc48155fb5224f4d6f0fb1c4b7eed6a79db4 | Update util.js (#4700)
Typo fix | [
{
"path": "src/shared/util.js",
"patch": "@@ -82,7 +82,7 @@ export function cached<F: Function> (fn: F): F {\n }\n \n /**\n- * Camelize a hyphen-delmited string.\n+ * Camelize a hyphen-delimited string.\n */\n const camelizeRE = /-(\\w)/g\n export const camelize = cached((str: string): string => {",
"a... | 2017-01-12T13:43:04 |
huggingface/transformers | 1ee3b288a62c9de658e8be117d869c2a9b835a7c | cad74496ca19c463a5fcc0b35ef4a1c9da2b8c4e | [`from_pretrained`] Small refactor `from_pretrained`: move around unrelated stuff (#41445)
* drafts
* up
* simplify modeling utils
* more simplifications
* type kwargs
* up
* move more accelerate related stuff
* safeguarding?
* nits
* remove func when func is NOPE
* more
* nits
* styling
* yups
* up
* u... | [
{
"path": "docs/source/en/main_classes/model.md",
"patch": "@@ -42,7 +42,3 @@ set this to `False`.\n ## Pushing to the Hub\n \n [[autodoc]] utils.PushToHubMixin\n-\n-## Sharded checkpoints\n-\n-[[autodoc]] modeling_utils.load_sharded_checkpoint",
"additions": 0,
"deletions": 4,
"language": "Mark... | 2025-10-13T14:33:32 |
ggml-org/llama.cpp | 84c8e305e8010a1a3d43bdd0a25f737ac67809a4 | 2aaf0a2a2056d75d0dd53ab8a181473760e6ab22 | Fix missing messages on sibling navigation (#16408)
* fix: resolve message disappearing issue when navigating between regenerated siblings by using current leaf nodes instead of cached sibling IDs
* chore: update webui build output
* chore: update webui build output | [
{
"path": "tools/server/webui/src/lib/stores/chat.svelte.ts",
"patch": "@@ -860,15 +860,18 @@ class ChatStore {\n \t\t\tthis.currentResponse = '';\n \n \t\t\ttry {\n-\t\t\t\tconst assistantMessage = await this.createAssistantMessage();\n+\t\t\t\tconst parentMessageId =\n+\t\t\t\t\tthis.activeMessages.length... | 2025-10-03T10:51:40 |
ollama/ollama | b58d5d16b093f88a51c48b9310acb4100943d23b | 24580df958e0ce0d1ef839c7c47ddd938456a6f3 | fix mkdir on windows | [
{
"path": "cmd/cmd.go",
"patch": "@@ -722,7 +722,7 @@ func initializeKeypair() error {\n \t\t\treturn err\n \t\t}\n \n-\t\terr = os.MkdirAll(path.Dir(privKeyPath), 0o700)\n+\t\terr = os.MkdirAll(filepath.Dir(privKeyPath), 0o755)\n \t\tif err != nil {\n \t\t\treturn fmt.Errorf(\"could not create directory %w... | 2023-09-19T16:36:30 |
denoland/deno | 76da1f08dfcb165400a470b39a36d9f66f0825bb | 9006b37ea61590ff3f55ee3acf1f3677ae57e55e | fix(ext/node): implement getValidStdio for child_process (#31850)
Implements the internal `getValidStdio` function used by Node.js's
child_process module to validate stdio configurations. This function:
- Validates stdio is a string ('ignore', 'pipe', 'inherit',
'overlapped') or array
- Expands arrays to at least 3... | [
{
"path": "ext/node/polyfills/internal/child_process.ts",
"patch": "@@ -41,7 +41,9 @@ import {\n AbortError,\n ERR_INVALID_ARG_TYPE,\n ERR_INVALID_ARG_VALUE,\n+ ERR_INVALID_SYNC_FORK_INPUT,\n ERR_IPC_CHANNEL_CLOSED,\n+ ERR_IPC_SYNC_FORK,\n ERR_UNKNOWN_SIGNAL,\n } from \"ext:deno_node/internal/er... | 2026-01-16T13:21:43 |
vuejs/vue | 7131bc48155fb5224f4d6f0fb1c4b7eed6a79db4 | b60182d996ed1e976ce1697a3017cb702ddc0920 | separate initRender and initial mounting (fix #4693) | [
{
"path": "src/core/instance/init.js",
"patch": "@@ -39,10 +39,13 @@ export function initMixin (Vue: Class<Component>) {\n vm._self = vm\n initLifecycle(vm)\n initEvents(vm)\n+ initRender(vm)\n callHook(vm, 'beforeCreate')\n initState(vm)\n callHook(vm, 'created')\n- initRender... | 2017-01-11T20:01:23 |
ggml-org/llama.cpp | 0e1f83855609d73beaf05d818640b6cfd39d287b | ad126479c25cf983a0f994a08ba0911cf49ed62b | vulkan: Fix FA coopmat1 invalid array indexing (#16365)
When computing sinks, the cm1 shader was looping r from 0 to Br rather than
to rows_per_thread. I must have copied this from the scalar path (where it is
correct), and somehow it wasn't causing failures on current drivers. | [
{
"path": "ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_cm1.comp",
"patch": "@@ -358,8 +358,8 @@ void main() {\n }\n \n if ((p.mask_n_head_log2 & SINK_ENABLE_BIT) != 0) {\n- [[unroll]] for (uint32_t r = 0; r < Br; ++r) {\n- float sink = perElemOpGetSink(r, 0u, ACC_TYPE(0), iq2);\... | 2025-10-03T09:52:46 |
ollama/ollama | 66003e1d05874703b425717bc39f112fcb88136a | c345053a8bf47d5ef8f1fe15d385108059209fba | subprocess improvements (#524)
* subprocess improvements
- increase start-up timeout
- when runner fails to start fail rather than timing out
- try runners in order rather than choosing 1 runner
- embed metal runner in metal dir rather than gpu
- refactor logging and error messages
* Update llama.go
* Upd... | [
{
"path": "llm/ggml.go",
"patch": "@@ -166,15 +166,15 @@ func (c *containerLORA) Decode(r io.Reader) (model, error) {\n }\n \n var (\n-\tggmlInit sync.Once\n-\tggmlRunnerPath string\n+\tggmlInit sync.Once\n+\tggmlRunners []ModelRunner // a slice of ModelRunners ordered by priority\n )\n \n-func ggm... | 2023-09-18T19:16:32 |
denoland/deno | 0f632fbe19664975093d32b014d84910a9051c71 | c356d9a621fccad5047497238039cde10613b278 | fix: don't ever set 1005 outbound close code (#31845)
1005 should never be sent on the wire. Next.js errors when receiving
this.
---------
Co-authored-by: Divy Srivastava <me@littledivy.com> | [
{
"path": "ext/websocket/lib.rs",
"patch": "@@ -721,7 +721,10 @@ pub async fn op_ws_close(\n const EMPTY_PAYLOAD: &[u8] = &[];\n \n let frame = reason\n- .map(|reason| Frame::close(code.unwrap_or(1005), reason.as_bytes()))\n+ .map(|reason| match code {\n+ Some(code) => Frame::close(code, &rea... | 2026-01-16T09:53:54 |
huggingface/transformers | cad74496ca19c463a5fcc0b35ef4a1c9da2b8c4e | 3813a8e3a1663993b3ec44c455cab8af1beca2b5 | [model] Add VideoLLaMA3 implementation (#40499)
* Add VideoLLaMA3 implementation
* Run style fix
* Switch to modular
* Fix config and smart_resize
* Fix
* Fix
* Fix style
* Fix
* Ruff fix
* Rename
* Rename
* Fix
* Clean
* Fix consistency
* Add doc
* Fix
* Fix
* Fix doc
* Update generated code
* remo... | [
{
"path": "docs/source/en/_toctree.yml",
"patch": "@@ -1188,6 +1188,8 @@\n title: TVP\n - local: model_doc/udop\n title: UDOP\n+ - local: model_doc/video_llama_3\n+ title: VideoLlama3\n - local: model_doc/video_llava\n title: VideoLlava\n - local: model_... | 2025-10-13T13:54:34 |
vuejs/vue | 92657249ddbb68c7b28c7e9d6fcc8e68cf8a151f | 38b30b4e605aeca41a8889f706c19d5d359a784a | Mark node with static props as static (#4662)
* fix special static attrs as dom prop
* refactor | [
{
"path": "flow/compiler.js",
"patch": "@@ -82,6 +82,7 @@ declare type ASTElement = {\n plain?: boolean;\n pre?: true;\n ns?: string;\n+ staticProps?: Array<string>;\n \n component?: string;\n inlineTemplate?: true;",
"additions": 1,
"deletions": 0,
"language": "JavaScript"
},
{
... | 2017-01-10T22:38:32 |
ggml-org/llama.cpp | 136bda78c5679b266362b39c58338fff46fd2592 | 5113efd34ceda709292de26c72716c49e024fb32 | webui : Fix messages payload sent to chat completions (#16402)
* fix: Include just the currently active message branches instead of all in chat completions request
* chore: Build webui static output
* chore: Formatting
* chore: update webui build output | [
{
"path": "tools/server/webui/src/lib/stores/chat.svelte.ts",
"patch": "@@ -550,7 +550,6 @@ class ChatStore {\n \t\t\t\tawait this.updateConversationName(this.activeConversation.id, title);\n \t\t\t}\n \n-\t\t\tconst allMessages = await DatabaseStore.getConversationMessages(this.activeConversation.id);\n \t... | 2025-10-03T07:11:34 |
denoland/deno | 04e93df2e7b3b00e8d0698f8b0f62223b6d3379f | 47ea9b2ad7e377964965a2205038dc454ca5cd0f | fix(audit): respect --level flag for exit code (#31721)
Closes #31702 | [
{
"path": "cli/tools/pm/audit.rs",
"patch": "@@ -333,18 +333,24 @@ mod npm {\n let minimal_severity =\n AdvisorySeverity::parse(&audit_flags.severity).unwrap();\n print_report(\n- vulns,\n+ &vulns,\n advisories,\n response.actions,\n minimal_severity,\n audit_... | 2026-01-15T10:00:48 |
huggingface/transformers | 66d8d7a07759c1a3138272244ca5f81834de9aa5 | d621be82861ad02670731f9c6e5b7391855d534c | Fixed typos and formatting (#34215)
#hacktoberfest | [
{
"path": "i18n/README_fr.md",
"patch": "@@ -227,7 +227,7 @@ Le modèle lui-même est un module [`nn.Module` PyTorch](https://pytorch.org/doc\n \n 1. Choisissez le bon framework pour chaque partie de la vie d'un modèle :\n - Entraînez des modèles de pointe en 3 lignes de code.\n- - Transférer un seul m... | 2025-10-13T13:38:06 |
vuejs/vue | 367c2e34a1d53c3122c5487bb39606a2e8bca380 | a36dc07222632e7c961eadf6556d1c98f49f4007 | [weex] Use weex.requireModule instead of weex.require (#4682)
* [weex] use weex.requireModule instead of weex.require
* [weex] fix the deprecated __weex_require_module__ | [
{
"path": "src/entries/weex-framework.js",
"patch": "@@ -91,7 +91,7 @@ export function createInstance (\n const weexInstanceVar = {\n config,\n document,\n- require: moduleGetter\n+ requireModule: moduleGetter\n }\n Object.freeze(weexInstanceVar)\n \n@@ -112,7 +112,7 @@ export function c... | 2017-01-10T16:02:55 |
ggml-org/llama.cpp | 5113efd34ceda709292de26c72716c49e024fb32 | d64c8104f090b27b1f99e8da5995ffcfa6b726e2 | fix: track viewportHeight via window.innerHeight to avoid unwanted scrolling (#16356)
Use <svelte:window bind:innerHeight> instead of manual resize listener
Co-authored-by: Aleksander Grygier <aleksander.grygier@gmail.com> | [
{
"path": "tools/server/webui/src/routes/+layout.svelte",
"patch": "@@ -25,6 +25,7 @@\n \tlet isNewChatMode = $derived(page.url.searchParams.get('new_chat') === 'true');\n \tlet showSidebarByDefault = $derived(activeMessages().length > 0 || isLoading());\n \tlet sidebarOpen = $state(false);\n+\tlet innerHei... | 2025-10-03T06:01:31 |
ollama/ollama | 08d7c2a94485b9203a8d35e1c6becc7bb98c905b | bc9573dcb15180fdb1ecda44160dc128b5003e70 | fix error on upload chunk | [
{
"path": "server/upload.go",
"patch": "@@ -78,6 +78,8 @@ func uploadBlobChunked(ctx context.Context, requestURL *url.URL, layer *Layer, r\n \t\t\t\tTotal: layer.Size,\n \t\t\t\tCompleted: int(offset),\n \t\t\t})\n+\n+\t\t\treturn err\n \t\t}\n \n \t\toffset += chunk",
"additions": 2,
"deletions... | 2023-09-15T22:59:30 |
denoland/deno | 2bf94f1b4b73a0af2ae1bf80ce2aeef3a01932d0 | 02272d9573a7dbab6d0a3578dfda4f3d91f7e70d | fix(ci): rename `trusted-signing-action` to `artifact-signing-action` (#31849)
The GitHub action `Azure/trusted-signing-action` has been renamed to
`Azure/artifact-signing-action`.
Ref: https://github.com/Azure/artifact-signing-action/issues/107 | [
{
"path": ".github/workflows/ci.generate.ts",
"patch": "@@ -850,7 +850,7 @@ const ci = {\n \"github.repository == 'denoland/deno' &&\",\n \"(github.ref == 'refs/heads/main' || startsWith(github.ref, 'refs/tags/'))\",\n ].join(\"\\n\"),\n- uses: \"azure/trusted-sign... | 2026-01-14T22:19:35 |
huggingface/transformers | cf1e9834ec7339f4c605ba96d9c4e5cf59594cad | 6c901bdc0e1b9c727392f69b2d5fc15ceebec304 | Restore cuda graphs to continuous batching (#41421)
* Type hints and small fixes
* Remove unusued params
* Made slice inputs the default
* ruffed
* Updated some var name and moved index slicing
* Logging arg in example
* Added some padding debug var and reformat out cg
* First working CG, fixe size
* Working f... | [
{
"path": "examples/pytorch/continuous_batching.py",
"patch": "@@ -26,22 +26,25 @@\n \n from transformers import AutoModelForCausalLM, AutoTokenizer\n from transformers.generation import GenerationConfig\n+from transformers.generation.continuous_batching.requests import logger\n \n \n # MODEL_ID = \"Qwen/Qw... | 2025-10-13T09:57:56 |
vuejs/vue | e8d72d254bba2649d79979050fa748b1a46abfca | c9fa2e674b520bb64359e18cf6a6455a7d05a6e7 | fix alligator.io link in backers.md | [
{
"path": "BACKERS.md",
"patch": "@@ -74,7 +74,7 @@ It could be you!\n <img width=\"240px\" src=\"http://i.imgur.com/MRIkKgp.png\">\n </a>\n \n-<a href=\"https://alligator.io\"></a>\n+<a href=\"https://alligator.io\">\n <img width=\"240px\" src=\"https://alligator.io/images/alligator-logo.svg\">\n </a>\... | 2017-01-05T18:46:13 |
ggml-org/llama.cpp | ef07a4090672a3438d7f64f197795d7dc1c18957 | 34fcc5a4ace8c69476ef2ea3857f39a60334acc4 | ggml webgpu: add support for soft_max, optimize rms_norm (#16357)
* Add inplace softmax
* Move rms_norm to split row approach
* Update debug for supports_op
* clean up debug statements
* Update tests/test-backend-ops.cpp
Co-authored-by: Georgi Gerganov <ggerganov@gmail.com>
---------
Co-authored-by: Georgi Gerg... | [
{
"path": "ggml/include/ggml.h",
"patch": "@@ -1630,6 +1630,13 @@ extern \"C\" {\n float scale,\n float max_bias);\n \n+ GGML_API struct ggml_tensor * ggml_soft_max_ext_inplace(\n+ struct ggml_context * ctx,\n+ struct ggml_tens... | 2025-10-02T18:00:31 |
huggingface/transformers | 6c901bdc0e1b9c727392f69b2d5fc15ceebec304 | 58f9e133139653a120f9c6e73eb0511cc560fba7 | [SAM] Fix typing hints (#41506)
fix | [
{
"path": "src/transformers/models/sam/processing_sam.py",
"patch": "@@ -30,13 +30,15 @@\n if is_torch_available():\n import torch\n \n+NestedList = list[Union[Optional[float | int], \"NestedList\"]]\n+\n \n class SamImagesKwargs(ImagesKwargs, total=False):\n segmentation_maps: Optional[ImageInput]\... | 2025-10-13T09:52:00 |
denoland/deno | 02272d9573a7dbab6d0a3578dfda4f3d91f7e70d | bd6bce3e5fab4aa555ee811d7e7211a54b57865e | fix(io): ignore `BrokenPipe` error in `op_print` (#31844) | [
{
"path": "ext/io/lib.rs",
"patch": "@@ -1199,9 +1199,11 @@ pub fn op_print(\n ) -> Result<(), JsErrorBox> {\n let rid = if is_err { 2 } else { 1 };\n FileResource::with_file(state, rid, move |file| {\n- file\n- .write_all_sync(msg.as_bytes())\n- .map_err(JsErrorBox::from_err)\n+ match f... | 2026-01-14T20:43:27 |
vuejs/vue | c9fa2e674b520bb64359e18cf6a6455a7d05a6e7 | 8be3aa48594359af80c1aed2c07f348dd83dbdf9 | fix todomvc example director.js cdn link | [
{
"path": "examples/todomvc/index.html",
"patch": "@@ -4,7 +4,7 @@\n <meta charset=\"utf-8\">\n <title>Vue.js • TodoMVC</title>\n <link rel=\"stylesheet\" href=\"https://unpkg.com/todomvc-app-css/index.css\">\n- <script src=\"https:unpkg.com/director/build/director.js\"></script>\n+ <scrip... | 2017-01-05T16:14:57 |
ollama/ollama | d028853879dbb5a2eba3f18776931efffaff508d | 949553db236cf78daa3bdfe1a966a61a8a01d22e | fix: add falcon.go | [
{
"path": "llm/falcon.go",
"patch": "@@ -0,0 +1,22 @@\n+package llm\n+\n+const ModelFamilyFalcon = \"falcon\"\n+\n+const (\n+\tfalconModelType7B = 32\n+\tfalconModelType40B = 60\n+\tfalconModelType180B = 80\n+)\n+\n+func falconModelType(numLayer uint32) string {\n+\tswitch numLayer {\n+\tcase 32:\n+\t\tr... | 2023-09-13T21:47:32 |
denoland/deno | dc8e397b09055f3b0986da51276a6d6117f12067 | 79a4ddb127436e87f65a1012cc96d6ba460d7991 | fix(node/child_process): ChildProcess constructor without args (#31838)
Refactors ChildProcess class to match Node.js behavior where the
constructor can be called without arguments, and spawn() is called
separately with options.
Fix `test-child-process-constructor.js` | [
{
"path": "ext/node/polyfills/child_process.ts",
"patch": "@@ -236,7 +236,10 @@ export function spawn(\n options = normalizeSpawnArguments(command, args, options);\n \n validateAbortSignal(options?.signal, \"options.signal\");\n- return new ChildProcess(command, args, options);\n+\n+ const child = new... | 2026-01-14T18:09:30 |
huggingface/transformers | 58f9e133139653a120f9c6e73eb0511cc560fba7 | eb282422516157801606c38b82a4898499f1e1f0 | Fixed Type-hints in function defintions (#41525)
* Explicitly annotate default None parameters as Optional
* make style.
* make style.
* Fixed check_copies.
* fix consistency. | [
{
"path": "src/transformers/models/detr/image_processing_detr.py",
"patch": "@@ -1724,7 +1724,7 @@ def get_ids_area(masks, scores, dedup=False):\n \n # inspired by https://github.com/facebookresearch/detr/blob/master/models/detr.py#L258\n def post_process_object_detection(\n- self, outputs, t... | 2025-10-13T09:48:37 |
ggml-org/llama.cpp | 34fcc5a4ace8c69476ef2ea3857f39a60334acc4 | 91a2a5655658bb9ab77894716b82fae7ecb4b4d1 | model : Apertus model implementation (#15852)
* First attempt
* No permute during convert (fixes qk tensors), proper norm application.
* RoPE = NeoX
* Coherence!
* Migrate xielu params from tensors to hyperparameters
* Simple CUDA kernel
* Revert stupid LLM refactorings
* Chat template support
* configchecker ... | [
{
"path": "common/chat-parser.cpp",
"patch": "@@ -75,6 +75,35 @@ bool common_chat_msg_parser::add_tool_calls(const json & arr) {\n }\n return true;\n }\n+\n+bool common_chat_msg_parser::add_tool_call_short_form(const json & tool_call) {\n+ if (!tool_call.is_object() || tool_call.size() != 1) {\n+... | 2025-10-02T17:43:22 |
vuejs/vue | e259fc306eece6be8014d2ace0258d9775971cd5 | a0c37a9072c9cb9952df410baf1178c007f30803 | fix #4344 #4445 #4511 (#4646)
* fiz regular expression
* fix regular expression | [
{
"path": "src/compiler/parser/text-parser.js",
"patch": "@@ -4,7 +4,7 @@ import { cached } from 'shared/util'\n import { parseFilters } from './filter-parser'\n \n const defaultTagRE = /\\{\\{((?:.|\\n)+?)\\}\\}/g\n-const regexEscapeRE = /[-.*+?^${}()|[\\]/\\\\]/g\n+const regexEscapeRE = /[-.*+?^${}()|[\\]... | 2017-01-05T16:15:53 |
ollama/ollama | 0c5a454361c57f300254971b109e5e4ec937ebd3 | 7dee25a07f6057a4afd42097357ffbdae0fdaacc | fix model type for 70b | [
{
"path": "llm/gguf.go",
"patch": "@@ -99,6 +99,12 @@ func (llm *ggufModel) ModelType() string {\n \tswitch llm.ModelFamily() {\n \tcase \"llama\":\n \t\tif blocks, ok := llm.kv[\"llama.block_count\"].(uint32); ok {\n+\t\t\theads, headsOK := llm.kv[\"llama.head_count\"].(uint32)\n+\t\t\theadKVs, headsKVsOK ... | 2023-09-12T17:52:57 |
huggingface/transformers | 65cb8fac6dba2fc6e2e530ed97e983f38cf5d95f | 3927ffed31e3c0d2929bf98bd05b7c61fcc48b62 | [Qwen3VL] fix: hidden_states in place modification error (#41535)
```
File "transformers/models/qwen3_vl_moe/modeling_qwen3_vl_moe.py", line 941, in forward
hidden_states = self._deepstack_process(
^^^^^^^^^^^^^^^^^^^^^^^^
File "transformers/models/qwen3_vl_moe/modeling_qwen3_vl_moe.py", li... | [
{
"path": "src/transformers/models/qwen3_omni_moe/modeling_qwen3_omni_moe.py",
"patch": "@@ -1694,7 +1694,8 @@ def _deepstack_process(self, hidden_states, visual_pos_masks, visual_embeds):\n visual_pos_masks = visual_pos_masks[..., 0]\n visual_pos_masks = visual_pos_masks.to(hidden_states.de... | 2025-10-13T08:50:14 |
ggml-org/llama.cpp | 72ee736c447be4ededb51ab97904b4d33c90c261 | f09aefaa84d7f4d5df3f400f67944b94fef5b795 | ci : fix ubuntu-latest-cmake-rpc (disable ccache) (#16388) | [
{
"path": ".github/workflows/build.yml",
"patch": "@@ -362,11 +362,11 @@ jobs:\n id: checkout\n uses: actions/checkout@v4\n \n- - name: ccache\n- uses: ggml-org/ccache-action@v1.2.16\n- with:\n- key: ubuntu-latest-cmake-rpc\n- evict-old-files: 1d\n+ ... | 2025-10-02T11:51:36 |
denoland/deno | d8a3f1bb90b4ab66ecdd571a471bf5e661119839 | 0b4cbe3bc8491787e3aab78db35cee2ab2590bba | fix(node/child_process): use correct syscall name in spawn error (#31836)
The error object from spawn() was incorrectly setting syscall to
"spawnSync <file>" instead of "spawn <file>". | [
{
"path": "ext/node/polyfills/internal/child_process.ts",
"patch": "@@ -420,7 +420,7 @@ export class ChildProcess extends EventEmitter {\n } catch (err) {\n let e = err;\n if (e instanceof Deno.errors.NotFound) {\n- e = _createSpawnSyncError(\"ENOENT\", command, args);\n+ e = _... | 2026-01-13T19:26:19 |
vuejs/vue | da955d1a4f268c9183907306506f663c46b40406 | d3768c0268c74fe123b5bec715e32dce556d4ac2 | Use 'click' event for checkbox and radio (fix #4620) (#4639)
* listen to click event for checkbox and radio.
* add test cases | [
{
"path": "src/platforms/web/compiler/directives/model.js",
"patch": "@@ -62,7 +62,7 @@ function genCheckboxModel (\n : `:_q(${value},${trueValueBinding})`\n )\n )\n- addHandler(el, 'change',\n+ addHandler(el, 'click',\n `var $$a=${value},` +\n '$$el=$event.target,' +\n ... | 2017-01-04T19:43:52 |
huggingface/transformers | 3927ffed31e3c0d2929bf98bd05b7c61fcc48b62 | 7164924a7e83f223a2bf2e104bef98eabe545091 | [testing] reduce runtime of `HunYuanMoEV1IntegrationTest:test_model_generation` (#41373)
* fix
* fix
* fix
---------
Co-authored-by: ydshieh <ydshieh@users.noreply.github.com> | [
{
"path": "tests/models/hunyuan_v1_moe/test_modeling_hunyuan_v1_moe.py",
"patch": "@@ -16,6 +16,7 @@\n import unittest\n \n import pytest\n+import torch\n from parameterized import parameterized\n \n from transformers import is_torch_available\n@@ -99,10 +100,12 @@ def tearDown(self):\n def test_model_g... | 2025-10-10T20:27:01 |
ollama/ollama | f59c4d03f789f745ffce6c4b68fcf218fac85435 | f221637053850d62e488501b726d2acfbf43c6b1 | fix ggml arm64 cuda build (#520) | [
{
"path": "llm/llama.cpp/generate_linux.go",
"patch": "@@ -6,9 +6,10 @@ package llm\n //go:generate -command git-apply git -C ggml apply\n //go:generate git-apply ../ggml_patch/0001-add-detokenize-endpoint.patch\n //go:generate git-apply ../ggml_patch/0002-34B-model-support.patch\n-//go:generate cmake -S gg... | 2023-09-12T21:06:48 |
ggml-org/llama.cpp | bbd32bc0384fbfcf07369617de58856b1e0e95a3 | 2be72c2b121ee99f33927149265ce6073ade9e59 | ci : fix clean-up of old logs (#16381) | [
{
"path": "ci/run.sh",
"patch": "@@ -34,9 +34,9 @@ mkdir -p \"$2\"\n OUT=$(realpath \"$1\")\n MNT=$(realpath \"$2\")\n \n-rm -f \"$OUT/*.log\"\n-rm -f \"$OUT/*.exit\"\n-rm -f \"$OUT/*.md\"\n+rm -f $OUT/*.log\n+rm -f $OUT/*.exit\n+rm -f $OUT/*.md\n \n sd=`dirname $0`\n cd $sd/../\n@@ -607,6 +607,7 @@ if [ -z... | 2025-10-02T07:35:43 |
vuejs/vue | aeb2efab4035dcdaab4afb3bfd9e9cef39fdd120 | d80eff8eb61e48ac184536a159a0e96c4613e7d6 | [weex] fix the removeChild logic for text node (#4626) | [
{
"path": "src/platforms/weex/runtime/node-ops.js",
"patch": "@@ -34,8 +34,8 @@ export function insertBefore (node, target, before) {\n }\n \n export function removeChild (node, child) {\n- if (node.nodeType === 3) {\n- node.parentNode.setAttr('value', '')\n+ if (child.nodeType === 3) {\n+ node.setA... | 2017-01-02T16:38:02 |
denoland/deno | 0b4cbe3bc8491787e3aab78db35cee2ab2590bba | 323f43a05d5d0146260a38c0beb71755ed66d422 | fix(node/child_process): reject null bytes in arguments (#31828)
Add validation to reject null bytes (\0) in child_process arguments,
matching Node.js behavior. Null bytes in command arguments can cause
security issues as they may truncate strings at the OS level.
Validates null bytes in:
- file/command argument
... | [
{
"path": "ext/node/polyfills/child_process.ts",
"patch": "@@ -22,6 +22,7 @@ import {\n type SpawnSyncOptions,\n type SpawnSyncResult,\n stdioStringToArray,\n+ validateNullByteNotInArg,\n } from \"ext:deno_node/internal/child_process.ts\";\n import {\n validateAbortSignal,\n@@ -73,6 +74,7 @@ export... | 2026-01-13T17:17:49 |
huggingface/transformers | 7164924a7e83f223a2bf2e104bef98eabe545091 | 26a5368c4429a7b59f133ed283b4cfccd8b50ef7 | Fix Latex typesetting in documentation (#41177)
Fix Latex typsetting in documentation
Signed-off-by: Yuanyuan Chen <cyyever@outlook.com> | [
{
"path": "docs/source/en/cache_explanation.md",
"patch": "@@ -41,13 +41,13 @@ $$\n \n The query (`Q`), key (`K`), and value (`V`) matrices are projections from the input embeddings of shape `(b, h, T, d_head)`.\n \n-For causal attention, the mask prevents the model from attending to future tokens. Once a t... | 2025-10-10T15:54:27 |
ollama/ollama | 7dee25a07f6057a4afd42097357ffbdae0fdaacc | f221637053850d62e488501b726d2acfbf43c6b1 | fix falcon decode
get model and file type from bin file | [
{
"path": "llm/ggml.go",
"patch": "@@ -8,54 +8,77 @@ import (\n \t\"sync\"\n )\n \n-type ModelFamily string\n-\n-const ModelFamilyUnknown ModelFamily = \"unknown\"\n-\n-type ModelType uint32\n+type GGML struct {\n+\tmagic uint32\n+\tcontainer\n+\tmodel\n+}\n \n const (\n-\tModelType3B ModelType = 26\n-\tMo... | 2023-09-12T17:01:20 |
ggml-org/llama.cpp | e95fec640f43623911a2cd5bda8b19b1898c530c | ded67b94446ef4f7fd988dbde7a12deef9870c13 | HIP: Disable ROCWMMA fattn on CDNA when compiled against ROCWMMA 2.0.0 (#16221)
* HIP: Disable ROCWMMA fatt on CDNA when compiled against ROCWMMA 2.0.0
rocwmma 2.0.0 includes a bug in the code fakeing fp16 accumulation on CDNA
* CUDA: Fix volta condition in ggml_cuda_should_use_wmma_fattn | [
{
"path": "ggml/CMakeLists.txt",
"patch": "@@ -209,7 +209,6 @@ option(GGML_HIP \"ggml: use HIP\"\n option(GGML_HIP_GRAPHS \"ggml: use HIP graph, experimental, slow\" OFF)\n option(GGML_HIP_NO_VMM \"ggml: do not try to use HIP VMM\... | 2025-10-01T21:09:25 |
denoland/deno | 7be740f03525f2153fdebdb16b5bd0ecb884506e | 41dca90e29442b7cbb93e139d166059f158e133a | fix(node/child_process): improve arg type validation for execFile and fork (#31829)
- Improves argument type validation in `execFile()` and `fork()` to
match Node.js behavior
- Fixes error code from `ERR_INVALID_ARG_VALUE` to
`ERR_INVALID_ARG_TYPE` in fork argument validation
- Adds explicit array rejection in options... | [
{
"path": "ext/node/polyfills/child_process.ts",
"patch": "@@ -33,7 +33,6 @@ import {\n ERR_CHILD_PROCESS_IPC_REQUIRED,\n ERR_CHILD_PROCESS_STDIO_MAXBUFFER,\n ERR_INVALID_ARG_TYPE,\n- ERR_INVALID_ARG_VALUE,\n ERR_OUT_OF_RANGE,\n genericNodeError,\n } from \"ext:deno_node/internal/errors.ts\";\n@@... | 2026-01-13T16:02:24 |
vuejs/vue | 05591f177a80cee02818c53dcf6659a3b486e633 | c1f68dd96eab2ccc634d0b813cef4cc13e8211d3 | [weex] fix the text node could not be removed properly (#4606) | [
{
"path": "src/platforms/weex/runtime/node-ops.js",
"patch": "@@ -34,6 +34,10 @@ export function insertBefore (node, target, before) {\n }\n \n export function removeChild (node, child) {\n+ if (node.nodeType === 3) {\n+ node.parentNode.setAttr('value', '')\n+ return\n+ }\n node.removeChild(child)... | 2016-12-31T00:36:37 |
ollama/ollama | 7d749cc787d1c98de836f6c8ce15b27b7505e678 | e7e91cd71c2b5f9440e995087c7d7524c2cfed1c | fix darwin build script | [
{
"path": "scripts/build_darwin.sh",
"patch": "@@ -8,7 +8,7 @@ GO_LDFLAGS=\"$GO_LDFLAGS -X github.com/jmorganca/ollama/server.mode=release\"\n # build universal binary\n GOARCH=arm64 go generate ./...\n GOARCH=arm64 go build -ldflags \"$GO_LDFLAGS\" -o dist/ollama-darwin-arm64\n-rm -rf llm/llama.cpp/ggml/bu... | 2023-09-11T20:31:46 |
huggingface/transformers | feca4f3de717071b9e83e7cf8d0360289f750000 | c6042a41699745532ecee05de32daaef3b847770 | remove `tpu_num_cores` (#41383)
* remove-tpu-num-cores
* fix
* let's remove it
* style
* Update examples/legacy/seq2seq/finetune_tpu.sh
Co-authored-by: Mohamed Mekkouri <93391238+MekkCyber@users.noreply.github.com>
---------
Co-authored-by: Mohamed Mekkouri <93391238+MekkCyber@users.noreply.github.com> | [
{
"path": "examples/legacy/seq2seq/finetune_tpu.sh",
"patch": "@@ -16,8 +16,8 @@ export TPU_NUM_CORES=8\n \n # the proper usage is documented in the README, you need to specify data_dir, output_dir and model_name_or_path\n # run ./finetune_tpu.sh --help to see all the possible options\n-python xla_spawn.py ... | 2025-10-10T13:53:28 |
ggml-org/llama.cpp | ded67b94446ef4f7fd988dbde7a12deef9870c13 | 1fe4e38cc20af058ed320bd46cac934991190056 | llama : parameter conversion and loading fixes for PLaMo2 variants (#16075)
* Fix to use hidden_size_per_head
* Fix num heads
* Fix array
* Fix loading weights
* Support old GGUF converted by the previous version of llama.cpp
* Update src/llama-model.cpp
Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.... | [
{
"path": "convert_hf_to_gguf.py",
"patch": "@@ -4250,7 +4250,8 @@ def set_gguf_parameters(self):\n # This logic matches modeling_plamo.py's is_mamba function\n mamba_step = hparams.get(\"mamba_step\", 2)\n mamba_enabled = hparams.get(\"mamba_enabled\", True)\n- mamba_layers =... | 2025-10-01T21:08:15 |
vuejs/vue | 2ee516dfc8c72cf48ca7c77863216d93de0b30eb | 299ecfc19fa0f59effef71d24686bd7eb70ecbab | ensure updated hook is called after children are updated as well (fix #4599) | [
{
"path": "src/core/instance/lifecycle.js",
"patch": "@@ -109,9 +109,8 @@ export function lifecycleMixin (Vue: Class<Component>) {\n if (vm.$vnode && vm.$parent && vm.$vnode === vm.$parent._vnode) {\n vm.$parent.$el = vm.$el\n }\n- if (vm._isMounted) {\n- callHook(vm, 'updated')\n- ... | 2016-12-29T21:19:44 |
denoland/deno | 41dca90e29442b7cbb93e139d166059f158e133a | db55e48fa39e982bf0b7829e02b22f2661efa090 | tests: remake inspector tests handling out of order (#31788)
This PR rewrite the inspector tests from Rust to TypeScript, and also
uses a better strategy for out of order communication, fixing flaky
tests. | [
{
"path": "tests/integration/inspector_tests.rs",
"patch": "@@ -1,1905 +0,0 @@\n-// Copyright 2018-2026 the Deno authors. MIT license.\n-\n-use std::io::BufRead;\n-use std::process::ChildStderr;\n-use std::time::Duration;\n-\n-use anyhow::Error as AnyError;\n-use anyhow::anyhow;\n-use bytes::Bytes;\n-use fa... | 2026-01-13T15:59:04 |
huggingface/transformers | 60f6ec438a4e77dea878b753affa9da75bc8db3e | f9f8bf5a1062ce0293cbd42be0126e17a15446e9 | Fix detectron2 import (#41510)
* fix
* fix
* typo | [
{
"path": "src/transformers/utils/import_utils.py",
"patch": "@@ -724,7 +724,15 @@ def is_datasets_available() -> bool:\n \n @lru_cache\n def is_detectron2_available() -> bool:\n- return _is_package_available(\"detectron2\")\n+ # We need this try/except block because otherwise after uninstalling the l... | 2025-10-10T11:33:47 |
ggml-org/llama.cpp | 764799279f801c696503bacca490b4358587d94c | 2a9b63383a448ec18c754dfdc6e95cb853940a52 | Conversation action dialogs as singletons from Chat Sidebar + apply conditional rendering for Actions Dropdown for Chat Conversation Items (#16369)
* fix: Render Conversation action dialogs as singletons from Chat Sidebar level
* chore: update webui build output
* fix: Render Actions Dropdown conditionally only when... | [
{
"path": "tools/server/webui/src/lib/components/app/chat/ChatSidebar/ChatSidebar.svelte",
"patch": "@@ -1,9 +1,12 @@\n <script lang=\"ts\">\n \timport { goto } from '$app/navigation';\n \timport { page } from '$app/state';\n-\timport { ChatSidebarConversationItem } from '$lib/components/app';\n+\timport { ... | 2025-10-01T16:18:10 |
vuejs/vue | 6b1755ae8c9f44a20046331f6a3dad31b37bf177 | e120d1459d921de3e961343ea9233a8c6c438a44 | adjust hydration strategy (fix #4560) | [
{
"path": "src/core/vdom/patch.js",
"patch": "@@ -14,7 +14,7 @@\n \n import config from '../config'\n import VNode from './vnode'\n-import { makeMap, isPrimitive, _toString, warn } from '../util/index'\n+import { makeMap, isPrimitive, warn } from '../util/index'\n import { activeInstance } from '../instance... | 2016-12-27T22:26:58 |
denoland/deno | 06e4a27450d8c5e5f6a3b1a44ce676aff88122f6 | 651a868d9adaeae8314a4e34f44f80aadba25a00 | fix: allow multiple --allow-all args (#31833)
Currently passing multiple `--allow-all` (or `-A`) flags causes a crash,
but due to historic reasons some tools like VSCode, unconditionally added
`--allow-all` argument when running a debugger. This commit changes flag parsing so that `deno`
accepts `--allow-all` multiple... | [
{
"path": "cli/args/flags.rs",
"patch": "@@ -4840,8 +4840,11 @@ fn allow_all_arg() -> Arg {\n .conflicts_with(\"allow-ffi\")\n .conflicts_with(\"allow-import\")\n .conflicts_with(\"permission-set\")\n- .action(ArgAction::SetTrue)\n .help(\"Allow all permissions\")\n+ // So situations l... | 2026-01-13T14:52:54 |
huggingface/transformers | f9f8bf5a1062ce0293cbd42be0126e17a15446e9 | b4067472aee9b566237091dbcd3659dd2ce92004 | Revert `local_rank` deletion and some cleaning (#41504)
* forgot those
* clean
* Fix
* merge
* fix
* fix | [
{
"path": "examples/legacy/seq2seq/seq2seq_trainer.py",
"patch": "@@ -144,7 +144,7 @@ def _get_train_sampler(self) -> Optional[torch.utils.data.Sampler]:\n \n return (\n RandomSampler(self.train_dataset)\n- if self.args.local_rank == -1\n+ if self.ar... | 2025-10-10T10:23:04 |
ollama/ollama | de227b620f8b719072b9823f3d9f42123ee2106b | 63def6ca4917af980c88c75a71a2a1955ee36586 | fix nil pointer dereference | [
{
"path": "server/images.go",
"patch": "@@ -1300,7 +1300,7 @@ func makeRequestWithRetry(ctx context.Context, method string, requestURL *url.UR\n }\n \n func makeRequest(ctx context.Context, method string, requestURL *url.URL, headers http.Header, body io.Reader, regOpts *RegistryOptions) (*http.Response, er... | 2023-09-08T00:24:31 |
vuejs/vue | f5f8453050c16310b3bd33f0978a0f727043e771 | 610fe3159218fb08567a7857ec05ef687c129646 | Recommend user to provide reproduction repository for ssr issue (#4573)
* recommend user to provide reproduction repository for ssr issue
* fix typo
* update style | [
{
"path": ".github/CONTRIBUTING.md",
"patch": "@@ -28,7 +28,7 @@ Hi! I’m really excited that you are interested in contributing to Vue.js. Befo\n \n - It is recommended that you make a JSFiddle/JSBin/Codepen to demonstrate your issue. You could start with [this template](http://jsfiddle.net/df4Lnuw6/) that ... | 2016-12-27T21:09:21 |
ggml-org/llama.cpp | b2ba81dbe07b6dbea9c96b13346c66973dede32c | bf6f3b3a1965d70e07ca94aab7b01268fe483e96 | ci : fix ccache key for ubuntu-cpu-cmake (#16355)
* fix ccache key for ubuntu-cpu-cmake
* set it for release as well [no ci] | [
{
"path": ".github/workflows/build.yml",
"patch": "@@ -207,7 +207,7 @@ jobs:\n - name: ccache\n uses: ggml-org/ccache-action@v1.2.16\n with:\n- key: ubuntu-cpu-cmake\n+ key: ubuntu-cpu-cmake-${{ matrix.build }}\n evict-old-files: 1d\n \n - name: Build ... | 2025-09-30T19:41:42 |
denoland/deno | 210b309d2537ea0690755648ba132f58118fcc0c | 67ab6beddb674f01773f75837411befce49b8918 | fix(ext/web): handle empty TextDecoder input when stream is true (#31691)
for:
https://github.com/denoland/deno/issues/31605#issuecomment-3677640653
fixes the issue where TextDecoder with stream: true and an empty chunk
would return incorrect results.
It resolves the chunking problem for legacy encodings like big5,
... | [
{
"path": "ext/web/08_text_encoding.js",
"patch": "@@ -119,6 +119,10 @@ class TextDecoder {\n stream = options.stream;\n }\n \n+ if (stream && input.length === 0) {\n+ return \"\";\n+ }\n+\n try {\n /** @type {ArrayBufferLike} */\n let buffer = input;",
"additions": ... | 2026-01-12T10:15:52 |
huggingface/transformers | b4067472aee9b566237091dbcd3659dd2ce92004 | bc529a3368e9942bcf42ebf02bdd5546ef39f9be | Bump to hfh 1.0.0.rc5 to fix test (#41508) | [
{
"path": "setup.py",
"patch": "@@ -114,7 +114,7 @@\n \"GitPython<3.1.19\",\n \"hf-doc-builder>=0.3.0\",\n \"hf_xet\",\n- \"huggingface-hub==1.0.0.rc4\",\n+ \"huggingface-hub==1.0.0.rc5\",\n \"importlib_metadata\",\n \"ipadic>=1.0.0,<2.0\",\n \"jinja2>=3.1.0\",",
"additions... | 2025-10-10T10:12:08 |
vuejs/vue | 4d9de6a2e3e0422189d6a1f5d6d583a5266a9569 | 0cbc96101d8b4e03bc87776bfd55eb1a6bf3cceb | do not use looseEqual for checkbox v-model if has no true-value binding (fix #4586) | [
{
"path": "src/platforms/web/compiler/directives/model.js",
"patch": "@@ -56,8 +56,11 @@ function genCheckboxModel (\n const falseValueBinding = getBindingAttr(el, 'false-value') || 'false'\n addProp(el, 'checked',\n `Array.isArray(${value})` +\n- `?_i(${value},${valueBinding})>-1` +\n- `:... | 2016-12-27T20:58:57 |
ggml-org/llama.cpp | 16b0ca0d2e63fc1a0a43795f8552f9cb61d9f7a5 | 8d78cd2613ccdeb3cc86f59bc8f9ddd31cfbd3ed | Chatapi ignore empty sampling (#16330)
* fix: skip empty sampling fields instead of coercing to 0 in chat API options
* chore: update webui build output | [
{
"path": "tools/server/webui/src/lib/stores/chat.svelte.ts",
"patch": "@@ -221,69 +221,66 @@ class ChatStore {\n \t */\n \tprivate getApiOptions(): Record<string, unknown> {\n \t\tconst currentConfig = config();\n+\t\tconst hasValue = (value: unknown): boolean =>\n+\t\t\tvalue !== undefined && value !== nu... | 2025-09-30T17:18:54 |
denoland/deno | 67ab6beddb674f01773f75837411befce49b8918 | 466ab1212c7b838bea154d23b2ed8711669770f9 | fix(ext/fetch): read file urls via vfs (#31814) | [
{
"path": "Cargo.lock",
"patch": "@@ -2060,6 +2060,7 @@ dependencies = [\n \"deno_core\",\n \"deno_error\",\n \"deno_fs\",\n+ \"deno_io\",\n \"deno_path_util\",\n \"deno_permissions\",\n \"deno_tls\",\n@@ -2083,7 +2084,6 @@ dependencies = [\n \"tokio\",\n \"tokio-rustls\",\n \"tokio-socks\",\n- \"t... | 2026-01-12T01:16:12 |
huggingface/transformers | f4487ec52170558d6f1559073e8ca97f737f6d37 | e8194fe84f6622ea06593a2a371382bda43749c1 | fix gemma3n case failure (#41426)
* fix gemma3n case failure
Signed-off-by: Yao, Matrix <matrix.yao@intel.com>
* fix style
Signed-off-by: Yao, Matrix <matrix.yao@intel.com>
* Update dependency_versions_table.py
* change the case argument passing way to make the case PASS,
generation_config way need re-visit
Sign... | [
{
"path": "tests/models/gemma3n/test_modeling_gemma3n.py",
"patch": "@@ -31,7 +31,6 @@\n Gemma3nAudioConfig,\n Gemma3nAudioFeatureExtractor,\n Gemma3nConfig,\n- GenerationConfig,\n StaticCache,\n is_torch_available,\n )\n@@ -740,7 +739,7 @@ def setUp(self):\n audio_ds = load_d... | 2025-10-10T09:15:27 |
ollama/ollama | bf146fb072b8dbf49efa2f874959ad978c48bf29 | f0f49435771352c4d1e432351675d10d9e23c099 | fix retry on unauthorized chunk | [
{
"path": "server/upload.go",
"patch": "@@ -66,31 +66,39 @@ func uploadBlobChunked(ctx context.Context, requestURL *url.URL, layer *Layer, r\n \n \t\tsectionReader := io.NewSectionReader(f, int64(offset), chunk)\n \t\tfor try := 0; try < MaxRetries; try++ {\n+\t\t\tch := make(chan error, 1)\n+\n \t\t\tr, w ... | 2023-09-07T19:01:50 |
vuejs/vue | 02e2d99e277a1ba1bd42e1b81b2273903fdb7fbc | 0bb2d4e2b621950f5d44ed83b7e9ce15e282db68 | Revert "ensure leave transitions and enter transitions are triggered in the same frame (fix #4510)"
This reverts commit 92ad0bd378d38f84691120b1caa654ad195ea724. | [
{
"path": "examples/move-animations/index.html",
"patch": "@@ -15,19 +15,17 @@\n border: 1px solid #666;\n box-sizing: border-box;\n }\n- /* 1. define transition property, duration and easing */\n .fade-move, .fade-enter-active, .fade-leave-active {\n transition: all... | 2016-12-27T19:11:45 |
ggml-org/llama.cpp | 8d78cd2613ccdeb3cc86f59bc8f9ddd31cfbd3ed | d1c84a662daa91be975863913a59975b11458141 | ggml webgpu: support for rope,div,sub,glu,scale,cont operators (#16187)
* Work on rope
* Simplify inplace operation generation and combine mul/add generation
* Work on rope variants
* implement neox rope
* rope complete
* Add sub,div,glu operators
* implement scale op
* Update cpy shader to handle cont/more typ... | [
{
"path": "ggml/include/ggml.h",
"patch": "@@ -237,6 +237,8 @@\n #define GGML_EXIT_SUCCESS 0\n #define GGML_EXIT_ABORTED 1\n \n+// TODO: convert to enum https://github.com/ggml-org/llama.cpp/pull/16187#discussion_r2388538726\n+#define GGML_ROPE_TYPE_NORMAL 0\n #define GGML_ROPE_TYPE_NEOX 2\n #define GGML_... | 2025-09-30T16:57:51 |
denoland/deno | 466ab1212c7b838bea154d23b2ed8711669770f9 | ac0aa340f17438884acd50477a415ceb7937dd72 | fix(ext/node): `DatabaseSync` garbage collection invalidating associated resources (#31737) | [
{
"path": "ext/node/ops/sqlite/database.rs",
"patch": "@@ -745,7 +745,7 @@ impl DatabaseSync {\n \n Ok(StatementSync {\n inner: stmt_cell,\n- db: Rc::downgrade(&self.conn),\n+ db: self.conn.clone(),\n statements: Rc::clone(&self.statements),\n ignore_next_sqlite_error: Rc::cl... | 2026-01-10T04:53:49 |
huggingface/transformers | e8194fe84f6622ea06593a2a371382bda43749c1 | 9556b36b2f5599e190ff6fea644c1fcbdfb2717b | Fix some tests (#41503)
* fix
* fix
* doc | [
{
"path": "docs/source/ar/llm_tutorial_optimization.md",
"patch": "@@ -472,7 +472,7 @@ for _ in range(5):\n next_token_id = torch.argmax(next_logits, dim=-1)\n \n print(\"shape of input_ids\", next_token_id.shape)\n- print(\"length of key-value cache\", len(past_key_values[0][0])) # past_key_values ar... | 2025-10-10T09:05:09 |
ollama/ollama | f0f49435771352c4d1e432351675d10d9e23c099 | 07b4074e7b2e404fb577e98c103421acd2029de1 | fix get auth token | [
{
"path": "server/auth.go",
"patch": "@@ -103,7 +103,7 @@ func getAuthToken(ctx context.Context, redirData AuthRedirect, regOpts *Registry\n \n \theaders := make(http.Header)\n \theaders.Set(\"Authorization\", sig)\n-\tresp, err := makeRequest(ctx, \"GET\", redirectURL, headers, nil, regOpts)\n+\tresp, err ... | 2023-09-07T18:49:36 |
ggml-org/llama.cpp | a014310374a16f9204f2bcc1b458fc1eda67e469 | 35fb82497ec6c5904b0adb7e1c881a76c1c692db | cuda : Enable CUDA Graph usage for Nemotron Nano v2 (NemotronH) (#16328)
* Fix Nemotron Nano v2 9B not executing as CUDA Graph on NVIDIA GPUs
* fix to ensure test-backend-ops check passes | [
{
"path": "ggml/src/ggml-cuda/cpy.cu",
"patch": "@@ -329,7 +329,11 @@ void ggml_cuda_cpy(ggml_backend_cuda_context & ctx, const ggml_tensor * src0, gg\n } else\n #endif // GGML_USE_MUSA && GGML_MUSA_MUDNN_COPY\n {\n- CUDA_CHECK(cudaMemcpyAsync(src1_ddc, src0_ddc, ggml_nbytes(src0)... | 2025-09-30T08:13:22 |
huggingface/transformers | 5aca530b342e24c3f0bf51f1ee7d67a6a5b376fb | 4f323369db3ca4b56120a75367e2a707056fb621 | [Parakeet] unnecessary warning & auto mapping (#41412)
* add parakeet to CONFIG_MAPPING_NAMES
* TOKENIZER_MAPPING_NAMES update
* fix auto tokenizer
* update
* fix | [
{
"path": "src/transformers/models/auto/tokenization_auto.py",
"patch": "@@ -509,7 +509,7 @@\n (\"owlv2\", (\"CLIPTokenizer\", \"CLIPTokenizerFast\" if is_tokenizers_available() else None)),\n (\"owlvit\", (\"CLIPTokenizer\", \"CLIPTokenizerFast\" if is_tokenizers_available() else None)),\n ... | 2025-10-10T09:00:15 |
denoland/deno | b6594a790986aa4518558f8d50667bddb4801378 | d63261e57aff4686d0e977af76b30435ec122b2e | fix(ext/node): emit deprecation warning for SlowBuffer (#31772)
Adds the DEP0030 deprecation warning when SlowBuffer() is called,
matching Node.js behavior.
This fixes the node_compat test parallel/test-buffer-slow.js. | [
{
"path": "ext/node/polyfills/internal/buffer.mjs",
"patch": "@@ -153,10 +153,13 @@ export const constants = {\n };\n \n let bufferWarningAlreadyEmitted = false;\n+let slowBufferWarningAlreadyEmitted = false;\n let nodeModulesCheckCounter = 0;\n const bufferWarning = \"Buffer() is deprecated due to security... | 2026-01-10T02:14:56 |
ggml-org/llama.cpp | f1eb1cb1eba042b2d583234e80f65e2e225d8995 | de41f2b7bffab7582944b213ce12b605f8cf6e0f | kleidiai : fix work size and threads sync for fp16 (#16246) | [
{
"path": "ggml/src/ggml-cpu/CMakeLists.txt",
"patch": "@@ -513,9 +513,9 @@ function(ggml_add_cpu_backend_variant_impl tag_name)\n \n # Fetch KleidiAI sources:\n include(FetchContent)\n- set(KLEIDIAI_COMMIT_TAG \"v1.13.0\")\n+ set(KLEIDIAI_COMMIT_TAG \"v1.14.0\")\n set(... | 2025-09-30T07:07:20 |
denoland/deno | d63261e57aff4686d0e977af76b30435ec122b2e | cb0f39271a081e5426f1909d8eee9499db2fe56b | fix: add more details when panicking getting dir path from deno.json (#31805) | [
{
"path": "libs/config/deno_json/mod.rs",
"patch": "@@ -1433,11 +1433,15 @@ impl ConfigFile {\n }\n \n pub fn dir_path(&self) -> PathBuf {\n- url_to_file_path(&self.specifier)\n- .unwrap()\n- .parent()\n- .unwrap()\n- .to_path_buf()\n+ let path = url_to_file_path(&self.specifie... | 2026-01-09T21:09:36 |
huggingface/transformers | 4f323369db3ca4b56120a75367e2a707056fb621 | f5f3457278be00e7ba8a1766690f4d9a3e2248fa | Fixed tiny incorrect imports in `glm4v` (#41483)
Fixed tiny import issue in glm4v | [
{
"path": "src/transformers/models/glm4v/modular_glm4v.py",
"patch": "@@ -54,8 +54,8 @@\n Qwen2_5_VLVisionBlock,\n )\n from ..qwen2_vl.processing_qwen2_vl import (\n- Qwen2_VLProcessor,\n- Qwen2_VLProcessorKwargs,\n+ Qwen2VLProcessor,\n+ Qwen2VLProcessorKwargs,\n )\n \n \n@@ -1526,7 +1526,7 ... | 2025-10-10T08:57:01 |
vuejs/vue | 0bb2d4e2b621950f5d44ed83b7e9ce15e282db68 | 2540c37c15eed42cbbcf576545406097ec46a6e3 | Revert "fix enter transition flicker regression (fix #4576)"
This reverts commit 89987e8e2ce31c071ea937152dbf967b6213d4ad. | [
{
"path": "src/platforms/web/runtime/modules/transition.js",
"patch": "@@ -112,9 +112,9 @@ export function enter (vnode: VNodeWithData, toggleDisplay: ?() => void) {\n beforeEnterHook && beforeEnterHook(el)\n if (expectsCSS) {\n addTransitionClass(el, startClass)\n- addTransitionClass(el, activeC... | 2016-12-27T19:11:40 |
ggml-org/llama.cpp | a74a0d69f34f52fa10d4f0a7ce749fb3490d0774 | 5f7e166cbf7b9ca928c7fad990098ef32358ac75 | tests: override test_set_rows::max_nmse_err to allow for occasional rounding differences (#16295)
* tests: override test_set_rows::max_nmse_err to allow for occasional rounding differences
* apply similar error bounds to test_cpy | [
{
"path": "tests/test-backend-ops.cpp",
"patch": "@@ -2140,6 +2140,27 @@ struct test_set_rows : public test_case {\n }\n }\n }\n+\n+ double max_nmse_err() override {\n+ if (type == GGML_TYPE_Q4_0 || type == GGML_TYPE_Q4_1 || type == GGML_TYPE_IQ4_NL ||\n+ type ==... | 2025-09-30T00:26:34 |
huggingface/transformers | b543679d0ec057cb51a1d0be7b86df0e78556763 | ac7777be1675a46a4355142f6f48bf641e069bfb | [kernels] Remove RWKV kernel finally ! (#41493)
* rm kernel
* fix style | [
{
"path": "src/transformers/kernels/rwkv/wkv_cuda.cu",
"patch": "@@ -1,187 +0,0 @@\n-#include <stdio.h>\n-#include <assert.h>\n-\n-#define MIN_VALUE (-1e38)\n-\n-template <typename F>\n-__global__ void kernel_forward(\n- const int B, const int T, const int C, const F *__restrict__ const _w, const F *__re... | 2025-10-10T08:32:05 |
denoland/deno | 024c73e1209aadf6c170af2be31d88dbb2f94da1 | 3865cd86d4e740249fad0391cddbac4e87ae242e | chore: fix node_compat_test workflow (#31798) | [
{
"path": ".github/workflows/node_compat_test.yml",
"patch": "@@ -22,6 +22,8 @@ jobs:\n uses: actions/checkout@v4\n with:\n submodules: true\n+ - name: Setup Rust\n+ uses: dsherret/rust-toolchain-file@v1\n - name: Setup Deno\n uses: denoland/setup-deno@v2\... | 2026-01-09T19:12:40 |
vuejs/vue | 2540c37c15eed42cbbcf576545406097ec46a6e3 | 7f25f730c1b82a80158d488b025f138155664921 | fix dynamic transition type for higher-order transition components (fix #4545) | [
{
"path": "src/platforms/web/runtime/components/transition.js",
"patch": "@@ -62,6 +62,10 @@ function hasParentTransition (vnode) {\n }\n }\n \n+function isSameChild (child, oldChild) {\n+ return oldChild.key === child.key && oldChild.tag === child.tag\n+}\n+\n export default {\n name: 'transition',\n ... | 2016-12-27T17:10:13 |
ollama/ollama | 83c6be1666e8ccf9055e8b7813064644f0a1ad69 | 1adfa67589c7306159965080c10cfef892a6bc99 | fix model manifests (#477) | [
{
"path": "server/modelpath.go",
"patch": "@@ -115,7 +115,7 @@ func GetManifestPath() (string, error) {\n \t}\n \n \tpath := filepath.Join(home, \".ollama\", \"models\", \"manifests\")\n-\tif err := os.MkdirAll(filepath.Dir(path), 0o755); err != nil {\n+\tif err := os.MkdirAll(path, 0o755); err != nil {\n \... | 2023-09-06T21:30:08 |
ggml-org/llama.cpp | 5f7e166cbf7b9ca928c7fad990098ef32358ac75 | d72f5f7ba260b546190338b0b76f2f152581424f | Fix thinking blocks with quotes + add handling `[THINK]...[/THINK]` blocks (#16326)
* fix: prevent reasoning blocks with quotes from being truncated
* chore: update webui build output
* feat: Improve thinking content parsing
* test: Adds ChatMessage component stories for different thinking blocks
* chore: update w... | [
{
"path": "tools/server/webui/src/lib/utils/thinking.ts",
"patch": "@@ -1,20 +1,20 @@\n /**\n- * Parses thinking content from a message that may contain <think> tags\n+ * Parses thinking content from a message that may contain <think> tags or [THINK] tags\n * Returns an object with thinking content and cle... | 2025-09-29T16:49:47 |
huggingface/transformers | ac7777be1675a46a4355142f6f48bf641e069bfb | 17c31a98ac7c728615eb934220c1a443be69a7c5 | fix bnb model loading (#41499) | [
{
"path": "src/transformers/quantizers/auto.py",
"patch": "@@ -126,6 +126,12 @@ def from_dict(cls, quantization_config_dict: dict):\n \"The model's quantization config from the arguments has no `quant_method` attribute. Make sure that the model has been correctly quantized\"\n )\... | 2025-10-10T08:27:29 |
denoland/deno | 5c9b552a5b921c02f862b59335242223307bbfc9 | be13de0a0677913afc4e8523ff9f76bf177e6fbe | fix(runtime): handle `options` parameter in `Deno.errors.*` constructors (#31510)
`Deno.errors.*` are just errors. However, before, a `Deno.errors.*`
constructor wouldn't use the `options` parameter in the call to the
constructor call to the parent class (`super()`). Now it does.
In other words, the following code us... | [
{
"path": "runtime/js/01_errors.js",
"patch": "@@ -5,155 +5,155 @@ const { BadResource, Interrupted, NotCapable } = core;\n const { Error } = primordials;\n \n class NotFound extends Error {\n- constructor(msg) {\n- super(msg);\n+ constructor(msg, opts) {\n+ super(msg, opts);\n this.name = \"Not... | 2026-01-09T16:48:37 |
vuejs/vue | d111f3f1fdcac8206a4bed59aa0227fc622e884b | 89987e8e2ce31c071ea937152dbf967b6213d4ad | typo fix (#4583)
- nomralization / normalzation → normalization | [
{
"path": "src/compiler/codegen/index.js",
"patch": "@@ -302,10 +302,10 @@ function genChildren (el: ASTElement, checkSkip?: boolean): string | void {\n }\n }\n \n-// determine the normalzation needed for the children array.\n+// determine the normalization needed for the children array.\n // 0: no normal... | 2016-12-27T15:17:24 |
ggml-org/llama.cpp | d72f5f7ba260b546190338b0b76f2f152581424f | b77e6c18e1a6fac5705ed95f03af5436d67484c1 | ci : add AMD runners and workflows (#16249)
* ci : add AMD runners and workflows
* ci : move AMD jobs to separate workflow
* cont : fix paths | [
{
"path": ".github/workflows/build-amd.yml",
"patch": "@@ -0,0 +1,52 @@\n+name: CI (AMD)\n+\n+on:\n+ workflow_dispatch: # allows manual triggering\n+ push:\n+ branches:\n+ - master\n+ paths: [\n+ '.github/workflows/build-amd.yml',\n+ '**/CMakeLists.txt',\n+ '**/.cmake',\n+ '... | 2025-09-29T14:51:48 |
ollama/ollama | 1adfa67589c7306159965080c10cfef892a6bc99 | 790d24eb7bd1b15192a8acd79b60e225aaa6688e | tighten up the error string for `ollama show` flags (#476) | [
{
"path": "cmd/cmd.go",
"patch": "@@ -281,9 +281,9 @@ func ShowHandler(cmd *cobra.Command, args []string) error {\n \t}\n \n \tif flagsSet > 1 {\n-\t\treturn errors.New(\"only one of 'license', 'modelfile', 'parameters', 'system', or 'template' can be set\")\n+\t\treturn errors.New(\"only one of '--license'... | 2023-09-06T20:38:49 |
huggingface/transformers | c20849bad1df6a557f9a62fa7fb2abf469201296 | 776eea8612458f60bbe684f38ea34f85fe890486 | [`CI`] Fix copies on main (#41486)
fix copies | [
{
"path": "src/transformers/models/apertus/configuration_apertus.py",
"patch": "@@ -177,13 +177,6 @@ def __init__(\n attention_dropout=0.0,\n **kwargs,\n ):\n- super().__init__(\n- pad_token_id=pad_token_id,\n- bos_token_id=bos_token_id,\n- eos_tok... | 2025-10-09T16:38:14 |
vuejs/vue | 89987e8e2ce31c071ea937152dbf967b6213d4ad | 66bf120f7aa7543cfe684741cba6f5b638f551c0 | fix enter transition flicker regression (fix #4576) | [
{
"path": "src/platforms/web/runtime/modules/transition.js",
"patch": "@@ -112,9 +112,9 @@ export function enter (vnode: VNodeWithData, toggleDisplay: ?() => void) {\n beforeEnterHook && beforeEnterHook(el)\n if (expectsCSS) {\n addTransitionClass(el, startClass)\n+ addTransitionClass(el, activeC... | 2016-12-27T15:16:45 |
denoland/deno | be13de0a0677913afc4e8523ff9f76bf177e6fbe | 23da05af3744773ec26cc0975b4c9ef228d2c2f0 | fix(ext/node): add tls.setDefaultCACertificates support (#31522)
Closes #31208
The `tls.setDefaultCACertificates` API was missing in `node:tls`,
causing `TypeError` in applications that rely on programmatically
overriding the global CA store (e.g., for custom root CAs or testing
environments). | [
{
"path": "ext/node/lib.rs",
"patch": "@@ -390,6 +390,7 @@ deno_core::extension!(deno_node,\n ops::process::op_node_process_setuid,\n ops::process::op_process_abort,\n ops::tls::op_get_root_certificates,\n+ ops::tls::op_set_default_ca_certificates,\n ops::tls::op_tls_peer_certificate,\n ... | 2026-01-09T15:35:52 |
ollama/ollama | 213ffdb54899cd82a4ecf911f217542a1b37d2fc | d42d88386affab04250f521c8e92392c72270ce5 | macos `amd64` compatibility fixes | [
{
"path": "llm/ggml_llama.go",
"patch": "@@ -45,6 +45,7 @@ func osPath(llamaPath string) string {\n \tif runtime.GOOS == \"windows\" {\n \t\treturn path.Join(llamaPath, \"Release\")\n \t}\n+\n \treturn llamaPath\n }\n \n@@ -68,7 +69,9 @@ func initGGML() {\n \t\tcase \"windows\":\n \t\t\tfiles = []string{\"s... | 2023-09-06T01:33:27 |
huggingface/transformers | 3839d5101338bd44d35c09df76da1ec1b21964e2 | 78f79ba5af6546ffa0d8c5aea5404533dcf70f7d | `report_to` default changed to "none" + cleaning deprecated env var (#41375)
* reporting
* fix
* fix | [
{
"path": "docs/source/en/tasks/knowledge_distillation_for_image_classification.md",
"patch": "@@ -109,7 +109,6 @@ training_args = TrainingArguments(\n output_dir=\"my-awesome-model\",\n num_train_epochs=30,\n fp16=True,\n- logging_dir=f\"{repo_name}/logs\",\n logging_strategy=\"epoch\",\... | 2025-10-09T16:28:48 |
ggml-org/llama.cpp | b77e6c18e1a6fac5705ed95f03af5436d67484c1 | 2ddd3f2356c313385fafc19a9f0ce678c8fe03ee | ggml: riscv: add riscv spacemit backend (#15288)
* ggml: add spacemit backend
Change-Id: I249bdc043485d815a9c351867137bc1e27cc2e23
* add new line at end of file
Change-Id: I889ed1c85fb45e62350ecde0c06f70450cadfbe2
* add riscv zba extension limit
Change-Id: I321eb200f859751727afe5cae13074dfce2bb0ce
* fixed for re... | [
{
"path": ".github/workflows/build-linux-cross.yml",
"patch": "@@ -253,3 +253,47 @@ jobs:\n -DCMAKE_FIND_ROOT_PATH_MODE_INCLUDE=BOTH\n \n cmake --build build --config Release -j $(nproc)\n+\n+ ubuntu-24-riscv64-cpu-spacemit-ime-cross:\n+ runs-on: ubuntu-24.04\n+\n+ ... | 2025-09-29T14:50:44 |
vuejs/vue | de220a635d0e712bbd24f951bd4121780af054b9 | 84b7b82031ac70467cedcec801fde55a605e4321 | ignore text nodes between v-if conditions (fix #4533) | [
{
"path": "src/compiler/parser/index.js",
"patch": "@@ -352,6 +352,23 @@ function processIfConditions (el, parent) {\n }\n }\n \n+function findPrevElement (children: Array<any>): ASTElement | void {\n+ let i = children.length\n+ while (i--) {\n+ if (children[i].type === 1) {\n+ return children[i... | 2016-12-23T04:40:10 |
denoland/deno | 23da05af3744773ec26cc0975b4c9ef228d2c2f0 | 8cd2dd38235ffb14e4c12043477e88c649721cc2 | fix(ext/fetch): support ipv6 destination host with http and socks proxy (#31683)
Alternative to #31514 with proper SOCKS support and tests. Closes
#31504. | [
{
"path": "ext/fetch/proxy.rs",
"patch": "@@ -603,12 +603,14 @@ where\n \n if is_https {\n tunnel(&mut io, &orig_dst, user_agent, auth).await?;\n+ let host = orig_dst.host().unwrap();\n+ let host = host\n+ .strip_prefix('[')\n+ ... | 2026-01-09T15:29:10 |
huggingface/transformers | b450d55a91828a0868ad4345b53316e49000c9db | 1a3a5f5289e13980642315277ea5d9986622ffc8 | Remove `past_index` (#41384)
* remove-tpu-num-cores
* fix
* rm past index
* Revert "fix"
This reverts commit 7608a6c059210957d3a77812e66178c8b79a9313.
* Revert "remove-tpu-num-cores"
This reverts commit ef08a51d71389849851518d67d8ad6c9ea8f04fc. | [
{
"path": "src/transformers/trainer.py",
"patch": "@@ -2434,10 +2434,6 @@ def _inner_training_loop(\n if hasattr(epoch_dataloader, \"set_epoch\"):\n epoch_dataloader.set_epoch(epoch)\n \n- # Reset the past mems state at the beginning of each epoch if necessary.\n- ... | 2025-10-09T16:06:46 |
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.