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 | 1fba72361e8e0e865d569f7cd15e5aa50b41ac9a | ea30d827697cfe548dd6691e8fc191b010e2433f | Add SAM3 to 🤗 Transformers (#42285)
* add native image only sam3
* remove sam3 source code from transformers
* init sam3 refactor
* add working refactored sam3
* remove dedicated config mask fuser
* Add support for batch inference and inference in any dtype
* improve processing for mixed batched inputs
* updat... | [
{
"path": "docs/source/en/_toctree.yml",
"patch": "@@ -826,6 +826,8 @@\n title: RT-DETRv2\n - local: model_doc/sam2\n title: SAM2\n+ - local: model_doc/sam3_tracker\n+ title: Sam3Tracker\n - local: model_doc/segformer\n title: SegFormer\n - local: model_... | 2025-11-19T16:18:49 |
ggml-org/llama.cpp | 8c32d9d96d9ae345a0150cae8572859e9aafea0b | 0874693b449a847de2d052f2afb5d0cbe9409f92 | server: explicitly set the function name in lambda (#17538)
As [1] explained, the real debug message will be like:
"res operator(): operator() : queue result stop"
Set the name explicitly, the message is easy for debugging:
"res operator(): recv : queue result stop"
The left "operator()" is generated by 'RES... | [
{
"path": "tools/server/server-queue.cpp",
"patch": "@@ -199,7 +199,7 @@ server_task_result_ptr server_response::recv(const std::unordered_set<int> & id_\n std::unique_lock<std::mutex> lock(mutex_results);\n condition_results.wait(lock, [&]{\n if (!running) {\n- RE... | 2025-11-29T17:43:29 |
denoland/deno | c095dd46fdd994c94b2f8332a68106b976a7fa6e | c25bfa70d1cb05bdfce6b8c0ab48c03344167463 | fix: format wasm stack traces correctly, preserve context in unhandled rejection, improve cppgc inheritance perf (#32293)
Fixes https://github.com/denoland/deno/issues/30135
Fixes https://github.com/denoland/deno/issues/32239 | [
{
"path": "Cargo.lock",
"patch": "@@ -1929,9 +1929,9 @@ dependencies = [\n \n [[package]]\n name = \"deno_core\"\n-version = \"0.385.0\"\n+version = \"0.386.0\"\n source = \"registry+https://github.com/rust-lang/crates.io-index\"\n-checksum = \"e348ea58ad12f6bee5407768b571a8af09756246a45c83dfd27a5aeeb338c5e... | 2026-02-23T19:02:57 |
ollama/ollama | 987c16b2f755c0ac7f3c3b03d4b875dffc96a551 | 0759d8996e10277f2879d99cca03fa2174d0ac3c | Report more information about GPUs in verbose mode
This adds additional calls to both CUDA and ROCm management libraries to
discover additional attributes about the GPU(s) detected in the system, and
wires up runtime verbosity selection. When users hit problems with GPUs we can
ask them to run with `OLLAMA_DEBUG=1 ol... | [
{
"path": "gpu/gpu.go",
"patch": "@@ -259,6 +259,7 @@ func FindGPULibs(baseLibName string, patterns []string) []string {\n \n func LoadCUDAMgmt(cudaLibPaths []string) *C.cuda_handle_t {\n \tvar resp C.cuda_init_resp_t\n+\tresp.ch.verbose = getVerboseState()\n \tfor _, libPath := range cudaLibPaths {\n \t\tl... | 2024-01-23T00:03:32 |
vuejs/vue | 844a540c647dfa93dc714540953524830dd3475a | 6f312d636c3d6049dc9e60007f88ea871b8e8173 | fix: do not use MutationObserver in IE11
MutationObserver is unreliable in IE11. fix #6466 | [
{
"path": "src/core/util/env.js",
"patch": "@@ -99,13 +99,13 @@ export const nextTick = (function () {\n // \"force\" the microtask queue to be flushed by adding an empty timer.\n if (isIOS) setTimeout(noop)\n }\n- } else if (typeof MutationObserver !== 'undefined' && (\n+ } else if (!isIE... | 2017-09-12T23:56:18 |
huggingface/transformers | ea30d827697cfe548dd6691e8fc191b010e2433f | a74be0dcb9f2975cea06b6ad0a82f539e7c5763d | Fix validation checks order in benchmark_v2 (#42280)
fix validation checks order | [
{
"path": "benchmark_v2/framework/benchmark_config.py",
"patch": "@@ -203,6 +203,8 @@ def adapt_configs(\n config[\"sequence_length\"] = seqlen\n config[\"num_tokens_to_generate\"] = ntok\n config[\"gpu_monitoring\"] = monitor\n+ # Remove the old name so it get... | 2025-11-19T15:56:29 |
ggml-org/llama.cpp | 0874693b449a847de2d052f2afb5d0cbe9409f92 | 7d2add51d8e3759020d70f2ff3a76b5795ff67bc | common : fix json schema with '\' in literals (#17307)
* Fix json schema with '\' in literals
* Add "literal string with escapes" test | [
{
"path": "common/json-schema-to-grammar.cpp",
"patch": "@@ -268,10 +268,10 @@ static bool is_reserved_name(const std::string & name) {\n }\n \n std::regex INVALID_RULE_CHARS_RE(\"[^a-zA-Z0-9-]+\");\n-std::regex GRAMMAR_LITERAL_ESCAPE_RE(\"[\\r\\n\\\"]\");\n+std::regex GRAMMAR_LITERAL_ESCAPE_RE(\"[\\r\\n\\\... | 2025-11-29T16:06:32 |
denoland/deno | 13457aff18e73222a2d39c0aece82825f3956c5f | 6998e65d6866348677e45d5a370361a11c519ed8 | fix(ext/node): enable test-zlib-invalid-input-memory and test-zlib-un… (#32284)
Fixes two previously disabled zlib Node.js compatibility tests.
- **`parallel/test-zlib-invalid-input-memory.js`** — enabled and
passing.
Root cause was a two-layer problem:
1. `AsyncResource` in `async_hooks.ts` did not call `emitIni... | [
{
"path": "ext/node/polyfills/async_hooks.ts",
"patch": "@@ -8,6 +8,8 @@ import { core, primordials } from \"ext:core/mod.js\";\n import { validateFunction } from \"ext:deno_node/internal/validators.mjs\";\n import {\n AsyncHook,\n+ emitDestroy as emitDestroyHook,\n+ emitInit,\n executionAsyncId as in... | 2026-02-23T16:16:52 |
vuejs/vue | 6f312d636c3d6049dc9e60007f88ea871b8e8173 | a977740816d131d6490ec73e3df2444f51841c42 | fix(v-model): fix input listener with modifier blocking v-model update
fix #6552 | [
{
"path": "src/core/vdom/helpers/update-listeners.js",
"patch": "@@ -5,18 +5,22 @@ import { cached, isUndef } from 'shared/util'\n \n const normalizeEvent = cached((name: string): {\n name: string,\n+ plain: boolean,\n once: boolean,\n capture: boolean,\n- passive: boolean\n+ passive: boolean,\n+ ... | 2017-09-12T23:29:34 |
huggingface/transformers | a74be0dcb9f2975cea06b6ad0a82f539e7c5763d | e5c8a06fdd8ed7766bc4ea8952dba8b6fca9367f | Fix accelerate integration (#42264)
* fix
* add sorting
* typo
* fix
* improve doc
* doc
* doc | [
{
"path": "src/transformers/integrations/accelerate.py",
"patch": "@@ -159,61 +159,6 @@ def wrapper(*args, **kwargs):\n setattr(torch, torch_function_name, old_torch_function)\n \n \n-def find_tied_parameters(model: \"nn.Module\", **kwargs):\n- \"\"\"\n- Find the tied parameters in a given... | 2025-11-19T14:54:29 |
ollama/ollama | 0f5b8433195c505a673b5687897bfa7379ab3d13 | ffaf52e1e9e107c2b6946a6593b780f941d0a337 | Refine Accelerate usage on mac
For old macs, accelerate seems to cause crashes, but for
AVX2 capable macs, it does not. | [
{
"path": "llm/generate/gen_darwin.sh",
"patch": "@@ -18,7 +18,7 @@ sign() {\n fi\n }\n \n-COMMON_DARWIN_DEFS=\"-DCMAKE_OSX_DEPLOYMENT_TARGET=11.0 -DCMAKE_SYSTEM_NAME=Darwin -DLLAMA_ACCELERATE=off\"\n+COMMON_DARWIN_DEFS=\"-DCMAKE_OSX_DEPLOYMENT_TARGET=11.0 -DCMAKE_SYSTEM_NAME=Darwin\"\n \n case \"${GOAR... | 2024-01-23T00:25:56 |
ggml-org/llama.cpp | 47a268ea5000fc0f05fc1c5cd0062efebfe84b92 | 59d8d4e96341eb54f362ac3d583ef522566e2a39 | Vulkan: MMVQ Integer Dot K-Quant and MUL_MAT_ID support (#16900)
* vulkan: split mul_mmq_funcs for mul_mat_vecq use
* add mxfp4 mmvq
* add q2_k mmvq
* add q3_k mmvq
* add q4_k and q5_k mmvq
* add q6_k mmvq
* handle 4x4 quants per mmvq thread
* enable MUL_MAT_ID mmvq support
* enable subgroup optimizations for ... | [
{
"path": "ggml/src/ggml-vulkan/ggml-vulkan.cpp",
"patch": "@@ -613,9 +613,10 @@ struct vk_device_struct {\n vk_pipeline pipeline_dequant[GGML_TYPE_COUNT];\n vk_pipeline pipeline_dequant_mul_mat_vec_f32_f32[DMMV_WG_SIZE_COUNT][GGML_TYPE_COUNT][mul_mat_vec_max_cols];\n vk_pipeline pipeline_dequan... | 2025-11-29T08:37:22 |
denoland/deno | 6998e65d6866348677e45d5a370361a11c519ed8 | 12df8252d604c690a974f8b75d6a3b4ae6ef298d | fix(ext/node): support URL in child_process.fork modulePath (#32268)
## Summary
- Adds support for passing `URL` objects as the `modulePath` argument to
`child_process.fork()`, matching Node.js behavior
- Converts URL to file path using `toPathIfFileURL()` before string
validation
- Adds unit test for `fork()` with a ... | [
{
"path": "ext/node/polyfills/child_process.ts",
"patch": "@@ -45,6 +45,7 @@ import {\n convertToValidSignal,\n kEmptyObject,\n } from \"ext:deno_node/internal/util.mjs\";\n+import { toPathIfFileURL } from \"ext:deno_node/internal/url.ts\";\n import { kNeedsNpmProcessState } from \"ext:deno_process/40_p... | 2026-02-23T16:16:27 |
vuejs/vue | 3f42562de766a90f35ffa9a9efd1234814ad390c | 1a19c418b1e71f39d2095d60de59fcd650719523 | chore: fix comment typo (#6573)
correct "a input" to "an input" | [
{
"path": "dist/vue.js",
"patch": "@@ -83,7 +83,7 @@ function toString (val) {\n }\n \n /**\n- * Convert a input value to a number for persistence.\n+ * Convert an input value to a number for persistence.\n * If the conversion fails, return original string.\n */\n function toNumber (val) {",
"addition... | 2017-09-12T13:10:39 |
huggingface/transformers | 6ccacf3ab99df836ef257f1fa431896f4d87a9f1 | 5804c1fa474f2568e359e3fb5800cb7b13fb01e2 | [testing] fix some cases in xpu (#42273)
fix tests/models/sam_hq/test_modeling_sam_hq.py::SamHQModelIntegrationTest::test_inference_mask_generation_no_point
fix tests/models/mpt/test_modeling_mpt.py::MptIntegrationTests::test_model_logits
Signed-off-by: Wang, Yi <yi.a.wang@intel.com> | [
{
"path": "tests/models/mpt/test_modeling_mpt.py",
"patch": "@@ -528,6 +528,7 @@ def test_generation_batched(self):\n for i, predicted_output in enumerate(decoded_outputs):\n self.assertEqual(predicted_output, expected_output[i])\n \n+ @require_deterministic_for_xpu\n def test_mod... | 2025-11-19T14:11:53 |
ggml-org/llama.cpp | 59d8d4e96341eb54f362ac3d583ef522566e2a39 | d82b7a7c1d73c0674698d9601b1bbb0200933f29 | vulkan: improve topk perf for large k, fix overflow in unit tests (#17582) | [
{
"path": "ggml/src/ggml-vulkan/ggml-vulkan.cpp",
"patch": "@@ -10239,7 +10239,9 @@ static void ggml_vk_topk(ggml_backend_vk_context * ctx, vk_context& subctx, cons\n \n // Prefer going as small as num_topk_pipelines - 3 for perf reasons.\n // But if K is larger, then we need a larger workgr... | 2025-11-29T07:39:57 |
vuejs/vue | b77a298cf3d20691a3b32eb8d7f62d2a67a430ae | f477ecdbbb1a86ffcf4508ab0f7351277a649a7e | chore: fixed typo in changelog section (#6549) | [
{
"path": "README.md",
"patch": "@@ -109,7 +109,7 @@ Please make sure to read the [Contributing Guide](https://github.com/vuejs/vue/b\n \n ## Changelog\n \n-Details changes for each release are documented in the [release notes](https://github.com/vuejs/vue/releases).\n+Detailed changes for each release are ... | 2017-09-08T13:27:19 |
huggingface/transformers | 15b73034d7ded1f47c4d684564b4ecd558b99c20 | f0d026f033388829a552eec57f5ad8dd8e105068 | fix a typo: pbd -> pdb (#42268) | [
{
"path": "src/transformers/models/esm/modeling_esmfold.py",
"patch": "@@ -2257,7 +2257,7 @@ def infer(\n \n @staticmethod\n def output_to_pdb(output: dict) -> list[str]:\n- \"\"\"Returns the pbd (file) string from the model given the model output.\"\"\"\n+ \"\"\"Returns the pdb (file)... | 2025-11-19T13:28:27 |
denoland/deno | 12df8252d604c690a974f8b75d6a3b4ae6ef298d | 52a2337fbc6230922576d7d8aeff5d6dd1fc0f3d | fix(ext/node): support ipv6 host in `node:http` (#32258)
Closes #21864. Full reproduction of the bug:
```ts
import { once } from "node:events";
import { request } from "node:http";
await using _server = Deno.serve({ hostname: "::1" }, () => new Response());
const req = request("http://[::1]:8000").end();
const [res]... | [
{
"path": "ext/node/polyfills/http.ts",
"patch": "@@ -98,8 +98,12 @@ import { methods as METHODS } from \"node:_http_common\";\n import { deprecate } from \"node:util\";\n \n const { internalRidSymbol } = core;\n-const { ArrayIsArray, StringPrototypeToLowerCase, SafeArrayIterator } =\n- primordials;\n+cons... | 2026-02-23T15:46:00 |
ggml-org/llama.cpp | d82b7a7c1d73c0674698d9601b1bbb0200933f29 | 03914c7ef826caf0b6371a6d1de270cda102b542 | gguf-py : fix passing non-native endian tensors (editor-gui and new-metadata) (#17553)
gguf_new_metadata.py reads data from reader.
Reader doesn't byteswap tensors to native endianness.
But writer does expect tensors in native endianness to convert them
into requested endianness.
There are two ways to fix this: updat... | [
{
"path": "gguf-py/gguf/gguf_writer.py",
"patch": "@@ -371,10 +371,13 @@ def add_tensor_info(\n \n def add_tensor(\n self, name: str, tensor: np.ndarray[Any, Any], raw_shape: Sequence[int] | None = None,\n- raw_dtype: GGMLQuantizationType | None = None,\n+ raw_dtype: GGMLQuantizati... | 2025-11-28T19:53:01 |
huggingface/transformers | f0d026f033388829a552eec57f5ad8dd8e105068 | 383018e7b930bf119026466a579d49e6f43eb4da | [testing] fix `cwm` (#42261)
* fix
* fix
* fix
* fix
---------
Co-authored-by: ydshieh <ydshieh@users.noreply.github.com> | [
{
"path": "tests/models/cwm/test_modeling_cwm.py",
"patch": "@@ -17,6 +17,7 @@\n from transformers import is_torch_available\n from transformers.testing_utils import (\n cleanup,\n+ require_read_token,\n require_torch,\n require_torch_accelerator,\n slow,\n@@ -85,6 +86,7 @@ class CwmModel... | 2025-11-19T13:25:26 |
ollama/ollama | 730dcfcc7ab3c1fc24417887533eee8c72e45a1c | 5f81a33f43edea71edfb3d045e140595caeaa226 | Refine debug logging for llm
This wires up logging in llama.cpp to always go to stderr, and also
turns up logging if OLLAMA_DEBUG is set. | [
{
"path": "llm/dyn_ext_server.go",
"patch": "@@ -136,6 +136,12 @@ func newDynExtServer(library, model string, adapters, projectors []string, opts\n \n \tsparams.n_threads = C.uint(opts.NumThread)\n \n+\tif debug := os.Getenv(\"OLLAMA_DEBUG\"); debug != \"\" {\n+\t\tsparams.verbose_logging = C.bool(true)\n+\... | 2024-01-22T20:26:49 |
ggml-org/llama.cpp | 3ce7a65c2f2529a8fc566b4aead53b088f7faec2 | e072b2052e9250395e4a28a28d37806342ac5db1 | server: fix: /metrics endpoint returning JSON-escaped Prometheus format (#17386)
* fix: /metrics endpoint returning JSON-escaped Prometheus format
* mod: remove string overload from ok() method | [
{
"path": "tools/server/server.cpp",
"patch": "@@ -2713,7 +2713,8 @@ struct server_routes {\n \n res->headers[\"Process-Start-Time-Unix\"] = std::to_string(res_task->t_start);\n res->content_type = \"text/plain; version=0.0.4\";\n- res->ok(prometheus.str());\n+ res->status = 20... | 2025-11-28T18:14:00 |
denoland/deno | 52a2337fbc6230922576d7d8aeff5d6dd1fc0f3d | 8cd0019f5b45b7a7219ae7412dacca5ebf932cae | fix(ext/node): add `openAsBlob` export to `node:fs` (#32261)
## Summary
- Implements `fs.openAsBlob()` which reads a file and returns a
`Promise<Blob>`, matching the Node.js API
- Adds the function to both named and default exports of `node:fs`
- Adds unit tests covering basic usage, the `type` option, and error
handl... | [
{
"path": "ext/node/polyfills/fs.ts",
"patch": "@@ -141,9 +141,19 @@ import {\n } from \"ext:deno_node/internal/fs/streams.mjs\";\n import {\n Dirent,\n+ getValidatedPath,\n toUnixTimestamp as _toUnixTimestamp,\n } from \"ext:deno_node/internal/fs/utils.mjs\";\n import { glob, globPromise, globSync } f... | 2026-02-23T15:38:02 |
vuejs/vue | 0f00f8fc2b83b964bb929b729a7c9e3675b52106 | f323719f20d2e2e8f61c60a07080bbfa8b7c0273 | fix(ssr): better handle v-html hydration
fix #6519 | [
{
"path": "src/core/vdom/patch.js",
"patch": "@@ -547,27 +547,46 @@ export function createPatchFunction (backend) {\n if (!elm.hasChildNodes()) {\n createChildren(vnode, children, insertedVnodeQueue)\n } else {\n- let childrenMatch = true\n- let childNode = elm.fi... | 2017-09-07T02:24:51 |
huggingface/transformers | 383018e7b930bf119026466a579d49e6f43eb4da | 95d75eb7b16edda3437aec315fadd32b40684f8a | Remove doc files of other langs for deleted models (#42276)
* fix
* fix
---------
Co-authored-by: ydshieh <ydshieh@users.noreply.github.com> | [
{
"path": "docs/source/ja/_toctree.yml",
"patch": "@@ -252,8 +252,6 @@\n title: Blenderbot Small\n - local: model_doc/bloom\n title: BLOOM\n- - local: model_doc/bort\n- title: BORT\n - local: model_doc/byt5\n title: ByT5\n - local: model_doc/camembert\n@... | 2025-11-19T13:23:53 |
ollama/ollama | 27a2d5af5452b82a5d1ab44853fd55a2c63865be | 5f81a33f43edea71edfb3d045e140595caeaa226 | Debug logging on init failure | [
{
"path": "llm/dyn_ext_server.go",
"patch": "@@ -142,7 +142,9 @@ func newDynExtServer(library, model string, adapters, projectors []string, opts\n \tC.dyn_llama_server_init(llm.s, &sparams, &initResp)\n \tif initResp.id < 0 {\n \t\tmutex.Unlock()\n-\t\treturn nil, extServerResponseToErr(initResp)\n+\t\terr ... | 2024-01-22T20:08:22 |
ggml-org/llama.cpp | e072b2052e9250395e4a28a28d37806342ac5db1 | c6f7a423c8c87748ef563a99d81c3b1b05cecff0 | ggml : add GGML_SCHED_NO_REALLOC option to disable reallocations in ggml_backend_sched (#17276)
* ggml : add GGML_SCHED_NO_REALLOC option to disable reallocations in ggml_backend_sched
Enabled in ggml-ci for testing.
* llama : update worst-case graph for unified cache
* ci : disable op offload in some tests
* fix s... | [
{
"path": "ci/run.sh",
"patch": "@@ -45,7 +45,7 @@ sd=`dirname $0`\n cd $sd/../\n SRC=`pwd`\n \n-CMAKE_EXTRA=\"-DLLAMA_FATAL_WARNINGS=ON -DLLAMA_CURL=ON\"\n+CMAKE_EXTRA=\"-DLLAMA_FATAL_WARNINGS=ON -DLLAMA_CURL=ON -DGGML_SCHED_NO_REALLOC=ON\"\n \n if [ ! -z ${GG_BUILD_METAL} ]; then\n CMAKE_EXTRA=\"${CMA... | 2025-11-28T15:33:23 |
huggingface/transformers | 6db2dd319c4b063836784bb4e75f664771956e6b | e2fb8d6062a05f69f976cf6e39618df6c31a3bfd | Audio Flamingo3 - fix attention masking (#42278)
* fix attention masking
* use `input_features_mask` for paddinf | [
{
"path": "src/transformers/models/audioflamingo3/modeling_audioflamingo3.py",
"patch": "@@ -30,7 +30,7 @@\n from ...activations import ACT2FN\n from ...cache_utils import Cache, EncoderDecoderCache\n from ...generation import GenerationMixin\n-from ...masking_utils import eager_mask, padding_mask_function\... | 2025-11-19T11:43:57 |
denoland/deno | 8cd0019f5b45b7a7219ae7412dacca5ebf932cae | 9e7fb63fe12f68e8e268756b14abce99dda0ae75 | fix: detect bound require reexports in cjs analysis (#32163)
Part of https://github.com/denoland/deno/issues/31347
Fixes https://github.com/denoland/deno/issues/27898
This update solves the module resolution. So the behaviour of stdio
output of `redocly` becomes the same as Node.js.
However, the problem of hanging i... | [
{
"path": "Cargo.lock",
"patch": "@@ -1780,9 +1780,9 @@ dependencies = [\n \n [[package]]\n name = \"deno_ast\"\n-version = \"0.53.0\"\n+version = \"0.53.1\"\n source = \"registry+https://github.com/rust-lang/crates.io-index\"\n-checksum = \"8ac5df28715465cff3f03294564f38c9cd3c24c4cfb965a32fd1327efa111b02\"... | 2026-02-23T15:27:18 |
vuejs/vue | 679cd1fef448989bf645313c391e4134ecd9f593 | 08bc7595fd57f8f52db83ae1b6bc9b7a33cdd4f9 | fix: support prop type checking for primitive wrapper objects (#6450)
close #6447 | [
{
"path": "src/core/util/props.js",
"patch": "@@ -145,7 +145,12 @@ function assertType (value: any, type: Function): {\n let valid\n const expectedType = getType(type)\n if (simpleCheckRE.test(expectedType)) {\n- valid = typeof value === expectedType.toLowerCase()\n+ const t = typeof value\n+ ... | 2017-09-05T20:23:53 |
ggml-org/llama.cpp | ddf9f94389a614ce005347f1c3f60ce477df1be1 | ff55414c42522adbeaa1bd9c52c0e9db16942484 | server : add Anthropic Messages API support (#17570)
* server : add Anthropic Messages API support
* remove -@pytest.mark.slow from tool calling/jinja tests
* server : remove unused code and slow/skip on test_anthropic_vision_base64_with_multimodal_model in test_anthropic_api.py
* server : removed redundant n field... | [
{
"path": "tools/server/README.md",
"patch": "@@ -7,6 +7,7 @@ Set of LLM REST APIs and a simple web front end to interact with llama.cpp.\n **Features:**\n * LLM inference of F16 and quantized models on GPU and CPU\n * [OpenAI API](https://github.com/openai/openai-openapi) compatible chat completions and ... | 2025-11-28T11:57:04 |
denoland/deno | 9e7fb63fe12f68e8e268756b14abce99dda0ae75 | 38f3ffe3630c2f796fb7e9d9e1de58da21047f60 | fix(ext/node): implement timeout and killSignal for spawn() (#32283)
Fixes the `timeout` and `killSignal` options for `child_process.spawn()`
(and `fork()`, which delegates to `spawn()`). Previously these options
were accepted but silently ignored.
The fix sets a `setTimeout` after spawning that kills the child with ... | [
{
"path": "ext/node/polyfills/child_process.ts",
"patch": "@@ -222,9 +222,27 @@ export function spawn(\n options = normalizeSpawnArguments(command, args, options);\n \n validateAbortSignal(options?.signal, \"options.signal\");\n+ validateTimeout(options?.timeout);\n \n const child = new ChildProcess(... | 2026-02-23T14:37:46 |
vuejs/vue | 38d52182bf8915628314e2aea7d2cc41ec39a0d6 | 3c216755f6eb656c6d864265a8dc7b51b3ae971b | fix(types): allow variadic plugin use (#6363)
fix #6357 | [
{
"path": "types/test/plugin-test.ts",
"patch": "@@ -17,3 +17,4 @@ const installer: PluginFunction<Option> = function(Vue, option) { }\n \n Vue.use(plugin, new Option);\n Vue.use(installer, new Option);\n+Vue.use(installer, new Option, new Option, new Option);",
"additions": 1,
"deletions": 0,
"... | 2017-09-05T20:02:38 |
huggingface/transformers | e2fb8d6062a05f69f976cf6e39618df6c31a3bfd | a5c903f877fda21e739027eed133e03162eb7712 | 🚨 Generalize `get_decoder()` for multimodal and delete redundant code 🔪 (#42156)
* update some models
* update the rest
* add helper for encoder
* delete encoder code from models
* fix copies
* fix some tests but VLM will fail
* add encider tests simialr to decoder
* no print
* fix overwritten models
* an... | [
{
"path": "src/transformers/modeling_utils.py",
"patch": "@@ -2205,26 +2205,78 @@ def disable_input_require_grads(self):\n \"\"\"\n self._require_grads_hook.remove()\n \n+ def get_encoder(self, modality: Optional[str] = None):\n+ \"\"\"\n+ Best-effort lookup of the *encoder*... | 2025-11-19T10:46:39 |
ollama/ollama | ec3764538d4755448b326400093608040219fd5f | fa8c990e58c05f100d728420a33feb45400223d8 | Probe GPUs before backend init
Detect potential error scenarios so we can fallback to CPU mode without
hitting asserts. | [
{
"path": "llm/ext_server/ext_server.cpp",
"patch": "@@ -3,6 +3,27 @@\n // Necessary evil since the server types are not defined in a header\n #include \"server.cpp\"\n \n+// Low level API access to verify GPU access\n+#if defined(GGML_USE_CUBLAS)\n+#if defined(GGML_USE_HIPBLAS)\n+#include <hip/hip_runtime.... | 2024-01-21T23:39:59 |
ggml-org/llama.cpp | ff55414c42522adbeaa1bd9c52c0e9db16942484 | 73955f7d2a3ce1f36d7ecc14495e08957b51d113 | model : Qwen3 Next (#16095)
* Qwen3 Next - cleaned up version
* Whitespaces and stuff
* Correct minor errors
* Update src/llama-model.cpp
Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com>
* Misc. fixes.
* Clean up code, add missing hybrid qualifier
* Did someone transpose the SOLVE_TRI result matri... | [
{
"path": "convert_hf_to_gguf.py",
"patch": "@@ -4183,6 +4183,36 @@ def set_vocab(self):\n super().set_vocab()\n \n \n+@ModelBase.register(\"Qwen3NextForCausalLM\")\n+class Qwen3NextModel(Qwen2MoeModel):\n+ model_arch = gguf.MODEL_ARCH.QWEN3NEXT\n+\n+ def set_gguf_parameters(self):\n+ s... | 2025-11-28T11:02:56 |
denoland/deno | 1c5a7189e57cd1d3d0cc3ce7a2f6725ac5ede3ea | 5ebbfc4a83fff688222857629af8942bccc3b18b | fix(audit): gracefully handle malformed response (#32234)
Fixes an issue where `deno audit` fails to deserialize npm registry
responses when the `actions` field is missing.
## Changes Made
- Added `#[serde(default)]` attribute to the `actions` field in
`AuditResponse` struct in `cli/tools/pm/audit.rs` - this makes t... | [
{
"path": "cli/tools/pm/audit.rs",
"patch": "@@ -603,6 +603,7 @@ mod npm {\n \n #[derive(Debug, Deserialize)]\n pub struct AuditResponse {\n+ #[serde(default)]\n pub actions: Vec<AuditAction>,\n pub advisories: HashMap<i32, AuditAdvisory>,\n pub metadata: AuditMetadata,\n@@ -920,3 +921,30... | 2026-02-22T02:32:21 |
huggingface/transformers | a5c903f877fda21e739027eed133e03162eb7712 | 67302b043ea6af670c1c7859bd763c4784edc7d3 | Fix looping in torch guard decorator (#42260)
* fix
* add
* fix
* switch loop order for perfs
* typo | [
{
"path": "src/transformers/initialization.py",
"patch": "@@ -162,6 +162,25 @@ def copy_(tensor: torch.Tensor, other: torch.Tensor) -> torch.Tensor:\n return tensor\n \n \n+# Here, we need to check several modules imported, and hot patch all of them, as sometimes torch does\n+# something like `from torc... | 2025-11-18T17:35:07 |
vuejs/vue | 3c216755f6eb656c6d864265a8dc7b51b3ae971b | a67b795b3661f6f223420339f3eeb07a913207de | fix(provide): provide should default to parentVal during merging (#6473)
fix #6436 | [
{
"path": "src/core/util/options.js",
"patch": "@@ -96,7 +96,7 @@ export function mergeDataOrFn (\n : childVal\n const defaultData = typeof parentVal === 'function'\n ? parentVal.call(vm)\n- : undefined\n+ : parentVal\n if (instanceData) {\n return mergeData... | 2017-09-05T20:01:50 |
ollama/ollama | c08dfaa23d90d5df044f44f1ba89fc1e0b727757 | 3b76e736aed78e19879c184b6d44570e0a150c36 | fix: remove overwritten model layers
if create overrides a manifest, first add the older manifest's layers to
the delete map so they can be cleaned up | [
{
"path": "server/images.go",
"patch": "@@ -412,6 +412,13 @@ func realpath(mfDir, from string) string {\n }\n \n func CreateModel(ctx context.Context, name, modelFileDir string, commands []parser.Command, fn func(resp api.ProgressResponse)) error {\n+\tdeleteMap := make(map[string]struct{})\n+\tif manifest,... | 2024-01-19T22:58:36 |
ggml-org/llama.cpp | 6783b11fb0889d68d0046176b4cc92ceee1961b0 | 909072abcfed4798f86b14c0a79df057a9e6ab47 | models : fix LFM2 tensors (#17548) | [
{
"path": "src/llama-arch.cpp",
"patch": "@@ -2237,7 +2237,7 @@ static const std::map<llm_arch, std::map<llm_tensor, const char *>> LLM_TENSOR_N\n { LLM_TENSOR_SHORTCONV_INPROJ, \"blk.%d.shortconv.in_proj\" },\n { LLM_TENSOR_SHORTCONV_OUTPROJ, \"blk.%d.shortconv.out_proj\" },\n ... | 2025-11-27T14:04:29 |
huggingface/transformers | 67302b043ea6af670c1c7859bd763c4784edc7d3 | 9f311047860d46367d51d2d5b280286b10ba9466 | Fix bnb for the weights refactor (#42043)
* small fix
* nits
* ish
* up
* rev
* fix more tie weights keys
* small fixes
* nit
* update
* fix and fix
* fix a test
* glubs
* current shitty changes
* ship validated ones
* more
* more update
* more
* more
* more
* mllama
* more up
* fix ernie
* fix x... | [
{
"path": "src/transformers/core_model_loading.py",
"patch": "@@ -46,22 +46,6 @@\n \n logger = logging.get_logger(__name__)\n \n-str_to_torch_dtype = {\n- \"BOOL\": torch.bool,\n- \"U8\": torch.uint8,\n- \"I8\": torch.int8,\n- \"I16\": torch.int16,\n- \"F16\": torch.float16,\n- \"BF16\": t... | 2025-11-18T17:28:42 |
denoland/deno | 5ebbfc4a83fff688222857629af8942bccc3b18b | 28d94f497698d9ad36ce2a49119d96466c9a59d1 | fix(lsp): qualify diagnostic source and message from tsgo (#32251) | [
{
"path": "cli/lsp/tsgo.rs",
"patch": "@@ -886,6 +886,16 @@ impl TsGoServerInner {\n }\n }\n \n+fn qualify_tsgo_diagnostic(diagnostic: &mut lsp::Diagnostic) {\n+ diagnostic.source = Some(\"deno-ts\".to_string());\n+ if let Some(lsp::NumberOrString::Number(code)) = &diagnostic.code {\n+ diagnostic.mes... | 2026-02-21T17:38:33 |
vuejs/vue | 89f0d29f2d541aa5a1ac9690258cd7c7ee576ef6 | 538ad20d8a37fe7ee2463ff20ac9557af70e0d33 | fix: inherit SVG ns on component root node (#6511)
fix #6506 | [
{
"path": "src/core/vdom/create-element.js",
"patch": "@@ -91,7 +91,7 @@ export function _createElement (\n let vnode, ns\n if (typeof tag === 'string') {\n let Ctor\n- ns = config.getTagNamespace(tag)\n+ ns = (context.$vnode && context.$vnode.ns) || config.getTagNamespace(tag)\n if (confi... | 2017-09-05T19:12:35 |
ggml-org/llama.cpp | 909072abcfed4798f86b14c0a79df057a9e6ab47 | cd8370b40890dd40ae91a1cc0206107cf78c9303 | cuda : fix UMA detection on discrete GPUs. (#17537) | [
{
"path": "ggml/src/ggml-cuda/ggml-cuda.cu",
"patch": "@@ -3837,7 +3837,7 @@ static void ggml_backend_cuda_device_get_memory(ggml_backend_dev_t dev, size_t *\n \n // Check if UMA is explicitly enabled via environment variable\n bool uma_env = getenv(\"GGML_CUDA_ENABLE_UNIFIED_MEMORY\") != nullptr;\n... | 2025-11-27T11:35:35 |
denoland/deno | 28d94f497698d9ad36ce2a49119d96466c9a59d1 | a29b955a86990be8795778a456b38cc09086575b | fix(install): support `--allow-scripts` with `deno install -g --compile ...` (#32249) | [
{
"path": "cli/args/flags.rs",
"patch": "@@ -6587,6 +6587,7 @@ fn install_parse(\n \n if compile {\n flags.type_check_mode = TypeCheckMode::Local;\n+ allow_scripts_arg_parse(flags, matches)?;\n }\n \n flags.subcommand =",
"additions": 1,
"deletions": 0,
"language": "Rust"
... | 2026-02-20T22:17:15 |
vuejs/vue | 538ad20d8a37fe7ee2463ff20ac9557af70e0d33 | a7444975343f7828004d90bfb0deeb98db0f46e7 | fix(directive): should invoke unbind & inserted on inner component root element change
fix #6513 | [
{
"path": "src/core/vdom/patch.js",
"patch": "@@ -670,6 +670,16 @@ export function createPatchFunction (backend) {\n for (let i = 0; i < cbs.create.length; ++i) {\n cbs.create[i](emptyNode, ancestor)\n }\n+ // #6513\n+ // invoke insert ho... | 2017-09-05T16:59:38 |
ggml-org/llama.cpp | cd8370b40890dd40ae91a1cc0206107cf78c9303 | d21a76ac38ca9d95be610f72b613c7455fa04103 | ggml-cpu: aarm64: q4_K repack gemm and gemv implementations (dotprod only) (#17494)
* Enabled q4_K_4x8 path
* Fixed generic Q4_K 8x4 implementation
* wip: dotprod gemm
* Working arm q4_K dotprod gemm
Signed-off-by: Alberto Cabrera <alberto.cabrera@liquid.ai>
* Undo acc rename
Signed-off-by: Alberto Cabrera <albe... | [
{
"path": "ggml/src/ggml-cpu/arch-fallback.h",
"patch": "@@ -33,23 +33,27 @@\n // repack.cpp\n #define ggml_quantize_mat_q8_0_4x4_generic ggml_quantize_mat_q8_0_4x4\n #define ggml_quantize_mat_q8_0_4x8_generic ggml_quantize_mat_q8_0_4x8\n+#define ggml_quantize_mat_q8_K_4x4_generic ggml_quantize_mat_q8_K_4x4... | 2025-11-27T11:25:14 |
denoland/deno | a29b955a86990be8795778a456b38cc09086575b | cb0c6d565c0d16916d292914d24237362c712013 | fix(ext/node): enable test-stdio-closed tests (#32237)
- Fix `child_process` shell command translation to handle shell
redirections (`1>&-`, `< file`), piped commands (`cmd1 | cmd2`), and
shell variable prefixes (`"$NODE"`, `${NODE}`). Previously shell
operators were passed to the CLI arg translator, causing parse fai... | [
{
"path": "ext/node/ops/node_cli_parser.rs",
"patch": "@@ -51,8 +51,13 @@ pub struct TranslatedArgs {\n fn translate_to_deno_args(\n parsed_args: ParseResult,\n script_in_npm_package: bool,\n+ wrap_eval: bool,\n ) -> TranslatedArgs {\n- let options = TranslateOptions::for_child_process();\n+ let opti... | 2026-02-20T21:36:26 |
vuejs/vue | a7444975343f7828004d90bfb0deeb98db0f46e7 | 2d75aba5e1a3f25aa30f90f95e603ba424cf1730 | fix: ensure outer bindings on nested HOC are properly re-applied on inner root element change | [
{
"path": "src/core/vdom/patch.js",
"patch": "@@ -660,14 +660,18 @@ export function createPatchFunction (backend) {\n // component root element replaced.\n // update parent placeholder node element, recursively\n let ancestor = vnode.parent\n+ const patchable = isPatch... | 2017-09-05T15:29:22 |
ollama/ollama | eb76f3e37993ab0c7d1c0d1c361245e1597690aa | d017e3d0a6a95922f9efe2625824e708a3db13f9 | Fix CPU-only build under Android Termux enviornment.
Update gpu.go initGPUHandles() to declare gpuHandles variable before
reading it. This resolves an "invalid memory address or nil pointer
dereference" error.
Update dyn_ext_server.c to avoid setting the RTLD_DEEPBIND flag under
__TERMUX__ (Android). | [
{
"path": "gpu/gpu.go",
"patch": "@@ -63,6 +63,7 @@ func initGPUHandles() {\n \n \t// TODO - if the ollama build is CPU only, don't do these checks as they're irrelevant and confusing\n \n+\tgpuHandles = &handles{nil, nil}\n \tvar cudaMgmtName string\n \tvar cudaMgmtPatterns []string\n \tvar rocmMgmtName st... | 2024-01-15T09:37:44 |
ggml-org/llama.cpp | 142df17c9c296c846131041283c69edd2db754d8 | e509411cf142807c947b53b340d2d5594ce38120 | vulkan: use a fixed 1KB buffer for the add_rms_fusion opt (#17514) | [
{
"path": "ggml/src/ggml-vulkan/ggml-vulkan.cpp",
"patch": "@@ -5289,7 +5289,8 @@ static void ggml_vk_init(ggml_backend_vk_context * ctx, size_t idx) {\n ctx->prealloc_size_x = 0;\n ctx->prealloc_size_y = 0;\n ctx->prealloc_size_split_k = 0;\n- ctx->prealloc_size_add_rms_partials = 0;\n+ /... | 2025-11-27T05:32:30 |
denoland/deno | 2e293c82032011d78c4ed8c17f405b0fd84c543b | fa9d25ab6682f8f9fe843451c23cd5edaf3b2a71 | fix: deno install --force --compile (#32242) | [
{
"path": "cli/tools/installer/mod.rs",
"patch": "@@ -972,10 +972,28 @@ async fn install_global_compiled(\n };\n \n let output_path = PathBuf::from(&output);\n- if output_path.is_file() && !install_flags_global.force {\n- return Err(anyhow!(\n- \"Existing installation found. Aborting (Use -f to... | 2026-02-20T20:38:26 |
vuejs/vue | 2d75aba5e1a3f25aa30f90f95e603ba424cf1730 | 230c6ae7822347b9b2a659503291e45fcc58fe41 | chore: fixed a small typo in CONTRIBUTING.md (#6517) | [
{
"path": ".github/CONTRIBUTING.md",
"patch": "@@ -49,7 +49,7 @@ The `setup` script links two git hooks:\n - `pre-commit`: runs ESLint on staged files.\n - `commit-msg`: validates commit message format (see below).\n \n-### Commiting Changes\n+### Committing Changes\n \n Commit messages should follow the [c... | 2017-09-05T15:02:03 |
huggingface/transformers | 9f311047860d46367d51d2d5b280286b10ba9466 | d372b827547e0414809daf120cea4ba525bf601d | delete already deprecated models (#42235)
* fix
* push deleted files
* fix
* fix
* fix
* fix
* fix
* fix
* fix
* fix
* fix
* fix
---------
Co-authored-by: ydshieh <ydshieh@users.noreply.github.com> | [
{
"path": "docs/source/en/_toctree.yml",
"patch": "@@ -420,8 +420,6 @@\n title: BLOOM\n - local: model_doc/blt\n title: BLT\n- - local: model_doc/bort\n- title: BORT\n - local: model_doc/byt5\n title: ByT5\n - local: model_doc/camembert\n@@ -476,8 +474,6... | 2025-11-18T14:53:38 |
ggml-org/llama.cpp | e509411cf142807c947b53b340d2d5594ce38120 | 7cba58bbeac0c262bde7d45adb452133f53cb56f | server: enable jinja by default, update docs (#17524)
* server: enable jinja by default, update docs
* fix tests | [
{
"path": "common/arg.cpp",
"patch": "@@ -694,6 +694,12 @@ static bool is_autoy(const std::string & value) {\n }\n \n common_params_context common_params_parser_init(common_params & params, llama_example ex, void(*print_usage)(int, char **)) {\n+ // default values specific to example\n+ // note: we pl... | 2025-11-27T00:02:50 |
denoland/deno | fa9d25ab6682f8f9fe843451c23cd5edaf3b2a71 | 498d6efc6c6564fa68be273d6b403f7ccc881f41 | feat(fmt): add --fail-fast flag to deno fmt --check (#31438)
Adds a `--fail-fast` flag to `deno fmt --check` that stops checking
files immediately after finding the first formatting error, matching the
behavior of `deno test --fail-fast`. | [
{
"path": "cli/args/flags.rs",
"patch": "@@ -269,6 +269,7 @@ pub struct EvalFlags {\n #[derive(Clone, Default, Debug, Eq, PartialEq)]\n pub struct FmtFlags {\n pub check: bool,\n+ pub fail_fast: bool,\n pub files: FileFlags,\n pub permit_no_files: bool,\n pub use_tabs: Option<bool>,\n@@ -3142,6 +31... | 2026-02-20T16:55:14 |
vuejs/vue | 230c6ae7822347b9b2a659503291e45fcc58fe41 | f76d16ed9507d4c2a90243ea3d77ccf00df29346 | fix(vdom): avoid diff de-opt when both head/tail are different
fix #6502 | [
{
"path": "src/core/vdom/patch.js",
"patch": "@@ -401,7 +401,9 @@ export function createPatchFunction (backend) {\n newStartVnode = newCh[++newStartIdx]\n } else {\n if (isUndef(oldKeyToIdx)) oldKeyToIdx = createKeyToOldIdx(oldCh, oldStartIdx, oldEndIdx)\n- idxInOld = isDef(newS... | 2017-09-04T16:50:57 |
ollama/ollama | aac9ab4db756b999c6c95b1159fc357a3b7c44b3 | 1f5b7ff976460ad5305ab31998554944825f1f5f | fix show handler | [
{
"path": "server/routes.go",
"patch": "@@ -630,11 +630,10 @@ func ShowModelHandler(c *gin.Context) {\n \t\treturn\n \t}\n \n-\tvar model string\n \tif req.Model != \"\" {\n-\t\tmodel = req.Model\n+\t\t// noop\n \t} else if req.Name != \"\" {\n-\t\tmodel = req.Name\n+\t\treq.Model = req.Name\n \t} else {\n ... | 2024-01-18T23:36:50 |
ggml-org/llama.cpp | 5449367b2125d63069fad8d2ca13d0c5ebb2f003 | 1d594c295c8863e9077feeb50305a5e19493e6ee | Fix chunks being too small with small matrix sizes (#17526) | [
{
"path": "ggml/src/ggml-cpu/repack.cpp",
"patch": "@@ -1731,12 +1731,13 @@ template <typename BLOC_TYPE, int64_t INTER_SIZE, int64_t NB_COLS, ggml_type PAR\n nchunk0 = (nr0 + min_chunk_size - 1) / min_chunk_size;\n }\n \n- if (nth == 1 || nchunk0 < nth || disable_chunking) {\n+ ... | 2025-11-26T21:14:54 |
huggingface/transformers | 1acbd0b3275067764495e794c6e81ba1f2317ce0 | c40b370bd01539ba9a05a35995a2fb5dc467f373 | Make tests run in less time by reducing `batch_size` (#42213)
* fix
* fix
* 1
* 2
* 2
* 2
* 2
* 2
* 2
* 2
* 2
* 2
* so many batch_size=13 --> batch_size=2
---------
Co-authored-by: ydshieh <ydshieh@users.noreply.github.com> | [
{
"path": ".circleci/create_circleci_config.py",
"patch": "@@ -187,6 +187,11 @@ def to_dict(self):\n # If it's done already, the files are inside the directory `/test_data/`.\n {\"run\": {\"name\": \"fetch hub objects before pytest\", \"command\": \"cp -r /test_data/* . 2>/dev/null |... | 2025-11-18T13:10:47 |
vuejs/vue | f76d16ed9507d4c2a90243ea3d77ccf00df29346 | 8fc6bc882708a725693b50a55f56cfde653ee48d | fix(vdom): Don't replace input for text-like type change (#6344)
fix #6313 | [
{
"path": "src/core/vdom/patch.js",
"patch": "@@ -6,8 +6,6 @@\n *\n * modified by Evan You (@yyx990803)\n *\n-\n-/*\n * Not type-checking this because this file is perf-critical and the cost\n * of making flow understand it is not worth it.\n */\n@@ -17,6 +15,7 @@ import config from '../config'\n impo... | 2017-09-01T22:49:39 |
ggml-org/llama.cpp | 1d594c295c8863e9077feeb50305a5e19493e6ee | eec1e33a9ed71b79422e39cc489719cf4f8e0777 | clip: (minicpmv) fix resampler kq_scale (#17516)
* debug:"solve minicpmv precision problem"
* “debug minicpmv”
* Apply suggestion from @ngxson
---------
Co-authored-by: Xuan-Son Nguyen <thichthat@gmail.com> | [
{
"path": "tools/mtmd/clip.cpp",
"patch": "@@ -1175,10 +1175,11 @@ struct clip_graph {\n cb(K, \"resampler_K\", -1);\n cb(V, \"resampler_V\", -1);\n \n+ float resampler_kq_scale = 1.0f/ sqrtf(float(d_head));\n embeddings = build_attn(\n model.mm... | 2025-11-26T20:44:07 |
huggingface/transformers | c40b370bd01539ba9a05a35995a2fb5dc467f373 | b1bdf9cb390f4d11f3e2eb89bd7c75c0cae8b9a1 | Allow VLMs to have a correct `base_model` (#41589)
* allow VLMs to have a correct `base_model`
* fix copies
* fix copies?
* empty commit
* fix copies
* nits after rebase
* fix copies
* add a test
* skip more tests
* fiix copies, ig have to do it in all PRs after rebase | [
{
"path": "src/transformers/models/aria/modeling_aria.py",
"patch": "@@ -596,7 +596,7 @@ def _init_weights(self, module):\n @auto_docstring\n class AriaPreTrainedModel(PreTrainedModel):\n config: AriaConfig\n- base_model_prefix = \"\"\n+ base_model_prefix = \"model\"\n supports_gradient_checkp... | 2025-11-18T10:25:39 |
vuejs/vue | 02f8b806768d70c589e646c384e592e93387b994 | 50257a58810a83b27316a793adf8d59a81ef3cf0 | fix(ssr): fix cachedEscape memory issue
close #6332 | [
{
"path": "src/platforms/web/server/modules/attrs.js",
"patch": "@@ -1,6 +1,6 @@\n /* @flow */\n \n-import { cachedEscape } from '../util'\n+import { escape } from '../util'\n \n import {\n isDef,\n@@ -50,7 +50,7 @@ export function renderAttr (key: string, value: string): string {\n } else if (isEnumera... | 2017-09-01T22:08:12 |
denoland/deno | 498d6efc6c6564fa68be273d6b403f7ccc881f41 | dde0741dc94351724169250acbc3901fed71b0ae | fix(ext/node): `util.parseEnv` and `process.loadEnvFile` compatibility (#32183) | [
{
"path": ".github/workflows/ci.yml",
"patch": "@@ -1347,7 +1347,7 @@ jobs:\n if: '!startsWith(github.ref, ''refs/tags/'')'\n env:\n CARGO_PROFILE_DEV_DEBUG: 0\n- run: cargo test --locked --lib -p deno -p denort -p node_shim -p deno_lib -p deno_snapshots -p deno_bundle_runti... | 2026-02-20T15:44:02 |
ollama/ollama | 96cfb626415cd811ab545c134bd0d16fa7aca044 | 598d6d5572337afac759abf1995950866d96d044 | fix: normalize name path before splitting | [
{
"path": "server/modelpath.go",
"patch": "@@ -46,6 +46,7 @@ func ParseModelPath(name string) ModelPath {\n \t\tname = after\n \t}\n \n+\tname = strings.ReplaceAll(name, string(os.PathSeparator), \"/\")\n \tparts := strings.Split(name, \"/\")\n \tswitch len(parts) {\n \tcase 3:",
"additions": 1,
"de... | 2024-01-17T00:48:05 |
ggml-org/llama.cpp | 879d673759181987300a29989478c0f36c97e97b | 6ab4e50d9c65e7d94ab7ce135114b52e484019ab | vulkan: Implement top-k (#17418)
* vulkan: Implement top-k
Each pass launches workgroups that each sort 2^N elements (where N is usually 7-10)
and discards all but the top K. Repeat until only K are left. And there's a fast
path when K==1 to just find the max value rather than sorting.
* fix pipeline selection
* vu... | [
{
"path": "ggml/src/ggml-vulkan/ggml-vulkan.cpp",
"patch": "@@ -409,6 +409,7 @@ enum shader_reduction_mode {\n // argsort pipelines for up to 1<<10 invocations per workgroup\n static constexpr uint32_t num_argsort_pipelines = 11;\n static constexpr uint32_t num_topk_moe_pipelines = 10;\n+static constexpr ui... | 2025-11-26T15:45:43 |
vuejs/vue | d6e6f1deb180a4f47e94496724623b9e6d8e08b3 | d77b95317cedae299605fb692e2c7c67796b17cb | fix(v-model): Allow using array value with array v-model in checkboxes (#6220)
fix #6219 | [
{
"path": "src/platforms/web/compiler/directives/model.js",
"patch": "@@ -93,7 +93,7 @@ function genCheckboxModel (\n 'if(Array.isArray($$a)){' +\n `var $$v=${number ? '_n(' + valueBinding + ')' : valueBinding},` +\n '$$i=_i($$a,$$v);' +\n- `if($$el.checked){$$i<0&&(${value}=$$a.con... | 2017-09-01T21:35:41 |
huggingface/transformers | b1bdf9cb390f4d11f3e2eb89bd7c75c0cae8b9a1 | cd416f3c5c6974b5afd211759624283777a45c6e | 🚨 Delete generation params from model config (#41695)
* i am so confused, too many circular dependencies. Delete and see what happens
* pop if exists
* fix a few tests
* fix loading generation params from model config
* oh no, revert this
* replace audios with audio in docs
* fix tests
* fix last test
* i am... | [
{
"path": "src/transformers/configuration_utils.py",
"patch": "@@ -302,10 +302,9 @@ def __init__(\n self.sep_token_id = sep_token_id\n self.decoder_start_token_id = decoder_start_token_id\n \n- # Retrocompatibility: Parameters for sequence generation. While we will keep the ability to... | 2025-11-18T10:20:24 |
ollama/ollama | d9bfb2f08fcb18cedc5a036216b770e24fd44006 | 598d6d5572337afac759abf1995950866d96d044 | install: pin fedora to max 37
repos for fedora 38 and newer do not exist as of this commit
```
$ dnf config-manager --add-repo https://developer.download.nvidia.com/compute/cuda/repos/fedora38/x86_64/cuda-fedora38.repo
Adding repo from: https://developer.download.nvidia.com/compute/cuda/repos/fedora38/x86_64/cuda-fed... | [
{
"path": "scripts/install.sh",
"patch": "@@ -231,8 +231,8 @@ if ! check_gpu nvidia-smi || [ -z \"$(nvidia-smi | grep -o \"CUDA Version: [0-9]*\\\n case $OS_NAME in\n centos|rhel) install_cuda_driver_yum 'rhel' $(echo $OS_VERSION | cut -d '.' -f 1) ;;\n rocky) install_cuda_driver_yum 'rh... | 2024-01-16T19:45:12 |
vuejs/vue | d77b95317cedae299605fb692e2c7c67796b17cb | 06741f326625e2db78d092e586923b97ba006906 | fix: handle special case for allowfullscreen on <embed>
close #6202 | [
{
"path": "src/platforms/web/runtime/modules/attrs.js",
"patch": "@@ -64,7 +64,12 @@ function setAttr (el: Element, key: string, value: any) {\n if (isFalsyAttrValue(value)) {\n el.removeAttribute(key)\n } else {\n- el.setAttribute(key, key)\n+ // technically allowfullscreen is a boo... | 2017-09-01T21:00:31 |
ggml-org/llama.cpp | e6923caaec6312ea374822634df9f7621f31fe74 | 3e18dba9fd599ace0c3f2bc94bd74e5398d2efb6 | ggml : fix ARM feature verification (#17519)
On arm64 with `cmake` version 3.31.6, the final feature verification fails:
-- ARM detected flags: -mcpu=neoverse-v2+crc+sve2-aes+sve2-sha3+nossbs
-- Performing Test GGML_MACHINE_SUPPORTS_dotprod
-- Performing Test GGML_MACHINE_SUPPORTS_dotprod - Success
--... | [
{
"path": "ggml/src/ggml-cpu/CMakeLists.txt",
"patch": "@@ -224,7 +224,8 @@ function(ggml_add_cpu_backend_variant_impl tag_name)\n \n include(CheckCXXSourceCompiles)\n set(CMAKE_REQUIRED_FLAGS_SAVE ${CMAKE_REQUIRED_FLAGS})\n- set(CMAKE_REQUIRED_FLAGS \"${ARCH_FLAGS}\")\n+ ... | 2025-11-26T13:14:41 |
huggingface/transformers | cd416f3c5c6974b5afd211759624283777a45c6e | 1742d1198ddce1eb3e34c7e86330b708fb9fd496 | Reduce timing on CircleCI - part 1 (Use @slow for IntegrationTests) (#42206)
* fix 1
* fix 2: bark
* fix 2: mamba
* fix 4: Speech2TextModelIntegrationTests
* fix 5: Aria
* fix 6: RTDetrModelIntegrationTest
* fix 7: PLBartBaseIntegrationTest
* fix 8: XLMRobertaModelIntegrationTest
* fix 9: TvpModelIntegrationTe... | [
{
"path": "tests/models/aria/test_modeling_aria.py",
"patch": "@@ -226,6 +226,7 @@ def test_training_gradient_checkpointing_use_reentrant_false(self):\n \n @unittest.skipIf(SKIP, reason=\"A10 doesn't have enough GPU memory for this tests\")\n @require_torch\n+@slow\n class AriaForConditionalGenerationIntegr... | 2025-11-18T09:16:59 |
vuejs/vue | 06741f326625e2db78d092e586923b97ba006906 | c371fbde9c1251bc225dddd682b32621a31f2548 | fix: computed properties should not be cached during SSR
ref: vuejs/vuex#877 | [
{
"path": "src/core/instance/state.js",
"patch": "@@ -23,6 +23,7 @@ import {\n nativeWatch,\n validateProp,\n isPlainObject,\n+ isServerRendering,\n isReservedAttribute\n } from '../util/index'\n \n@@ -169,6 +170,8 @@ const computedWatcherOptions = { lazy: true }\n function initComputed (vm: Compon... | 2017-09-01T16:51:29 |
ggml-org/llama.cpp | 05872ac88532585fb198f1113be43ac654955cba | 55ab25caf5caec6840d502325c8480b8afccc0b4 | convert : fix big-endian conversion (#17431)
* Fix convert_hf_to_gguf.py script on s390x
Assume converted model data is originally little-endian.
Byteswap data on s390x after reading it to put values in correct presentation
for any transformation needed, like calculating weight tensors.
Then byteswap data to little-... | [
{
"path": "convert_hf_to_gguf.py",
"patch": "@@ -10061,6 +10061,25 @@ class LazyTorchTensor(gguf.LazyBase):\n torch.uint8: np.uint8,\n }\n \n+ # only used when byteswapping data. Only correct size is needed\n+ _dtype_byteswap_map: dict[torch.dtype, type] = {\n+ torch.float64: np.flo... | 2025-11-25T13:18:16 |
ollama/ollama | eef50accb425a1815625f496bc5b24d5211d5610 | 05d53de7a10fd95fc475c27b0235b8b51601c2ba | Fix show parameters (#2017) | [
{
"path": "server/routes.go",
"patch": "@@ -15,7 +15,6 @@ import (\n \t\"path/filepath\"\n \t\"reflect\"\n \t\"runtime\"\n-\t\"strconv\"\n \t\"strings\"\n \t\"sync\"\n \t\"syscall\"\n@@ -668,27 +667,12 @@ func GetModelInfo(req api.ShowRequest) (*api.ShowResponse, error) {\n \tcs := 30\n \tfor k, v := range ... | 2024-01-16T18:34:44 |
huggingface/transformers | 1742d1198ddce1eb3e34c7e86330b708fb9fd496 | 16924cd33ad57c19a290121ca83f4ceefab8f1b7 | [loading] Fix device when source and target are different (#42246)
* fix device
* fix
* CI
* simplify a bit | [
{
"path": "src/transformers/core_model_loading.py",
"patch": "@@ -312,16 +312,16 @@ class ConversionEntry:\n GLOBAL_WORKERS = min(16, (os.cpu_count() or 8) * 2) # NVMe: 8-16; HDD/NFS: 2-4\n \n \n-def _materialize_copy(tensor, dtype=None):\n+def _materialize_copy(tensor, device=None, dtype=None):\n tens... | 2025-11-18T08:57:21 |
ggml-org/llama.cpp | 134e6940caf5c64071b7f3b7bc6c2f32f1b3a5a4 | 0543f928a3ae576e6e16d3bbf02c0bf9fddba688 | llama : skip output reordering for single token batches (#17466)
This commit adds a check to skip the output reordering logic when
n_outputs == 1. With a single output token, the data is trivially
sorted and the reordering code is currently doing unnecessary work
(resetting and rebuilding output_ids to the same values... | [
{
"path": "src/llama-context.cpp",
"patch": "@@ -1248,7 +1248,7 @@ int llama_context::decode(const llama_batch & batch_inp) {\n \n // make the outputs have the same order they had in the user-provided batch\n // note: this is mostly relevant for recurrent models atm\n- if (!sorted_out... | 2025-11-24T20:06:17 |
vuejs/vue | 0529040c17b8632032a43d142aac88386f6b4a1f | 172dbf9faf4cb71dff72c77fdfe80fa1932d1ba3 | fix: deep clone slot vnodes on re-render
fix #6372 | [
{
"path": "src/core/instance/render.js",
"patch": "@@ -78,9 +78,13 @@ export function renderMixin (Vue: Class<Component>) {\n } = vm.$options\n \n if (vm._isMounted) {\n- // clone slot nodes on re-renders\n+ // if the parent didn't update, the slot nodes will be the ones from\n+ // la... | 2017-08-29T22:47:10 |
huggingface/transformers | 16924cd33ad57c19a290121ca83f4ceefab8f1b7 | 266d3b05684ed5968c613f00b208909f03c72c0e | Stop inheriting tests (again) (#42247)
* Stop inheriting tests!
* Just use a del instead
* fixup
* Stop using del!
* make fixup | [
{
"path": "tests/models/cohere2/test_modeling_cohere2.py",
"patch": "@@ -34,8 +34,7 @@\n torch_device,\n )\n \n-from ...models.cohere.test_modeling_cohere import CohereModelTest, CohereModelTester\n-from ...test_configuration_common import ConfigTester\n+from ...models.cohere.test_modeling_cohere import... | 2025-11-18T08:31:11 |
denoland/deno | 08d2450c90cef5ca8abe52e389fee6e5221ca159 | ff153750f2b240b910b56cf36aea8c94cc603c00 | fix(ci): use noble apt repo for LLVM 21 packages (#32241)
The CI was configured to install LLVM 21 from the jammy (Ubuntu 22.04)
apt repository, but the runners use Ubuntu 24.04 (Noble). The mismatched
distro codename caused 404 errors when downloading packages. | [
{
"path": ".github/workflows/ci.generate.ts",
"patch": "@@ -117,8 +117,8 @@ sudo apt-get -qq remove \\\n 'clang-12*' 'clang-13*' 'clang-14*' 'clang-15*' 'clang-16*' 'clang-17*' 'clang-18*' 'clang-19*' 'llvm-12*' 'llvm-13*' 'llvm-14*' 'llvm-15*' 'llvm-16*' 'llvm-17*' 'llvm-18*' 'llvm-19*' 'lld-12*' 'lld-13... | 2026-02-20T04:20:28 |
ggml-org/llama.cpp | 0543f928a3ae576e6e16d3bbf02c0bf9fddba688 | b61de2b2df4ff07e6d6de96320fb311d96908b7a | HIP: WMMA-MMQ kernels for RDNA 4 (#17156)
* first commit naive test to enable mmq for RDNA4
* adding appropriate WMMA instructions
* git rebase on top of master: fixing the correctness of the mat mul operations, updating layout mappings for RDNA4
* clean up merge conflicts
* add comments and code clean up
* PR cl... | [
{
"path": "ggml/src/ggml-cuda/mma.cuh",
"patch": "@@ -73,34 +73,7 @@ namespace ggml_cuda_mma {\n static constexpr int I = I_;\n static constexpr int J = J_;\n \n-#if defined(GGML_USE_HIP)\n-#if defined(RDNA4)\n- static constexpr int ne = I * J / 32;\n- T x[ne] = {0};\n-\n- ... | 2025-11-24T19:00:10 |
vuejs/vue | 172dbf9faf4cb71dff72c77fdfe80fa1932d1ba3 | f4ee63ca1059efcbdcacea9f6c1ce4e9433e2baf | fix(ssr): should also escape static text content
fix #6345 | [
{
"path": "src/server/optimizing-compiler/codegen.js",
"patch": "@@ -22,8 +22,8 @@ import {\n applyModelTransform\n } from './modules'\n \n+import { escape } from 'web/server/util'\n import { optimizability } from './optimizer'\n-\n import type { CodegenResult } from 'compiler/codegen/index'\n \n export t... | 2017-08-29T21:32:28 |
huggingface/transformers | 266d3b05684ed5968c613f00b208909f03c72c0e | 8d6c4583bfa58230188fae074c91945e014e73c1 | Fix UnboundLocalError in RT-DETR loss computation (#42224)
* Fix UnboundLocalError in RT-DETR loss computation
Initialize auxiliary_outputs to None before conditional use to prevent
UnboundLocalError when config.auxiliary_loss is False.
Fixes the error:
UnboundLocalError: local variable 'auxiliary_outputs' referen... | [
{
"path": "src/transformers/loss/loss_rt_detr.py",
"patch": "@@ -447,6 +447,7 @@ def RTDetrForObjectDetectionLoss(\n outputs_loss = {}\n outputs_loss[\"logits\"] = logits\n outputs_loss[\"pred_boxes\"] = pred_boxes\n+ auxiliary_outputs = None\n if config.auxiliary_loss:\n if denoi... | 2025-11-17T17:06:10 |
denoland/deno | ff153750f2b240b910b56cf36aea8c94cc603c00 | 964288509b72a5f4217f61bb85e104f41ef09c24 | fix(lsp): typescript-go basic resolution (#32016) | [
{
"path": "cli/lsp/documents.rs",
"patch": "@@ -1296,6 +1296,49 @@ impl DocumentModules {\n module\n }\n \n+ pub fn module_for_tsgo_referrer(\n+ &self,\n+ uri: &Uri,\n+ compiler_options_key: &CompilerOptionsKey,\n+ ) -> Option<Arc<DocumentModule>> {\n+ let document = self.documents.get(u... | 2026-02-19T17:50:21 |
ggml-org/llama.cpp | 6ab8eacddf50cda653b1e27521bd88945c41df1b | 2d50b9d8cb6b6c0ef935809af61ad4958be47648 | examples : add -kvu to batched usage example [no ci] (#17469)
This commit adds the --kv-unified flag to the usage example
in the README.md file for the batched example.
The motivation for this is that without this flag the example will fail
with the following error:
```console
Hello my name is
split_equal: sequential... | [
{
"path": "examples/batched/README.md",
"patch": "@@ -3,7 +3,7 @@\n The example demonstrates batched generation from a given prompt\n \n ```bash\n-./llama-batched -m ./models/llama-7b-v2/ggml-model-f16.gguf -p \"Hello my name is\" -np 4\n+./llama-batched -m ./models/llama-7b-v2/ggml-model-f16.gguf -p \"Hell... | 2025-11-24T13:38:45 |
ollama/ollama | 3ca5f69ce889c4ba16086fbcfb388c4c940aa421 | 2ecb2472769ba0bb364aeeb109659750ad8153ce | Fix typo in arm mac arch script | [
{
"path": "llm/generate/gen_darwin.sh",
"patch": "@@ -18,7 +18,7 @@ case \"${GOARCH}\" in\n ;;\n \"arm64\")\n CMAKE_DEFS=\"-DCMAKE_SYSTEM_PROCESSOR=arm64 -DCMAKE_OSX_ARCHITECTURES=arm64 -DLLAMA_METAL=on ${CMAKE_DEFS}\"\n- ARHC=\"arm64\"\n+ ARCH=\"arm64\"\n ;;\n *)\n echo \"GOARCH must ... | 2024-01-14T16:32:57 |
huggingface/transformers | 8d6c4583bfa58230188fae074c91945e014e73c1 | 2cc9152da08dc9a2d1ac5e738b8eb764ca369717 | Fix processor test for glm (#42233)
fix arg name | [
{
"path": "tests/models/glm46v/test_processor_glm46v.py",
"patch": "@@ -215,7 +215,7 @@ def test_apply_chat_template_video_frame_sampling(self):\n add_generation_prompt=True,\n tokenize=True,\n return_dict=True,\n- video_fps=video_fps,\n+ fps=video_f... | 2025-11-17T16:36:30 |
vuejs/vue | 59dbd4a414394a3ce581f9fbd9554da9af9e4b1d | a43d66743be2bd62b2398090663e41eeaf0dc75f | fix: ensure $attrs and $listeners are always objects (#6441)
fix #6263 | [
{
"path": "flow/component.js",
"patch": "@@ -29,8 +29,8 @@ declare interface Component {\n $slots: { [key: string]: Array<VNode> };\n $scopedSlots: { [key: string]: () => VNodeChildren };\n $vnode: VNode; // the placeholder node for the component in parent's render tree\n- $attrs: ?{ [key: string] : ... | 2017-08-29T20:59:39 |
denoland/deno | 964288509b72a5f4217f61bb85e104f41ef09c24 | e64ffe6a4f61e0ddeaaff6fb5b4cea6b0264c45f | fix(ext/node): handle shell redirections in child_process exec (#32087)
When child_process.exec() runs a shell command like "deno" "script.js" <
"input.txt", transformDenoShellCommand failed to add run -A because
shell redirections (<, >, |) were included in the args passed to the CLI
parser, triggering the metacharac... | [
{
"path": "ext/node/lib.rs",
"patch": "@@ -291,6 +291,7 @@ deno_core::extension!(deno_node,\n ops::process::op_process_abort,\n ops::process::op_node_process_constrained_memory<TSys>,\n ops::node_cli_parser::op_node_translate_cli_args,\n+ ops::shell::op_node_parse_shell_args,\n ops::tls::... | 2026-02-19T17:48:40 |
ggml-org/llama.cpp | 697edfeead9769d68387dec6884e0ecac23d2e4e | dbb852b549adf29609ec53b518f7922a982f14b9 | ggml : remove dirty flag from version string (ggml/1391)
This commit removes the "-dirty" suffix from the GGML version string.
The motivation for this change is to ensure that the version string
works with different ways of checking out ggml and using it in projects.
By removing the dirty flag from the version string... | [
{
"path": "ggml/CMakeLists.txt",
"patch": "@@ -25,16 +25,17 @@ if(GIT_EXE)\n )\n endif()\n \n-# Build the version string with optional dirty flag\n set(GGML_VERSION \"${GGML_VERSION_BASE}\")\n-if(GGML_GIT_DIRTY AND NOT GGML_GIT_DIRTY EQUAL 0)\n- set(GGML_VERSION \"${GGML_VERSION}-dirty\")\n-endif()\n... | 2025-11-24T11:51:50 |
huggingface/transformers | 2cc9152da08dc9a2d1ac5e738b8eb764ca369717 | 8637f6e7ae1e9c901cae7f1d2db9d33d63af72a2 | FIX Broken PEFT adapter loading (#42187)
* FIX Broken PEFT adapter loading
For some time now, loading PEFT adapters directly with transformers is
broken when using revisions or subfolders.
To check, run:
RUN_SLOW=1 pytest tests/peft_integration/test_peft_integration.py -k
test_peft_from_pretrained_hub_kwargs
This ... | [
{
"path": "src/transformers/integrations/peft.py",
"patch": "@@ -654,17 +654,18 @@ def maybe_load_adapters(\n token_from_adapter_kwargs = adapter_kwargs.pop(\"token\", None)\n \n if _adapter_model_path is None:\n+ peft_kwargs = adapter_kwargs.copy()\n+ for arg_name in (\"cache_dir\", \... | 2025-11-17T15:02:57 |
vuejs/vue | a43d66743be2bd62b2398090663e41eeaf0dc75f | d4d553ced75d8c73e75b85cec398be4b09f6f669 | fix(transition): consider async placeholder as valid child to return (#6369)
fix #6256 | [
{
"path": "src/core/vdom/helpers/get-first-component-child.js",
"patch": "@@ -1,12 +1,13 @@\n /* @flow */\n \n import { isDef } from 'shared/util'\n+import { isAsyncPlaceholder } from './is-async-placeholder'\n \n export function getFirstComponentChild (children: ?Array<VNode>): ?VNode {\n if (Array.isArr... | 2017-08-29T20:09:20 |
denoland/deno | 8e15a1f450b91aa44a6f6b9881c9be717575e750 | 3239a3fe2d1da2cd93df4cbd099c8d0973699fa0 | feat: add `navigator.platform` (#30795)
Fixes https://github.com/denoland/deno/issues/24117
---------
Signed-off-by: Bedis Nbiba <bedisnbiba@gmail.com>
Co-authored-by: Bartek Iwańczuk <biwanczuk@gmail.com> | [
{
"path": "cli/tsc/dts/lib.deno.window.d.ts",
"patch": "@@ -362,6 +362,7 @@ interface Navigator {\n readonly userAgent: string;\n readonly language: string;\n readonly languages: string[];\n+ readonly platform: string;\n }\n \n /**",
"additions": 1,
"deletions": 0,
"language": "Unknown"
... | 2026-02-19T08:19:17 |
ollama/ollama | 2ecb2472769ba0bb364aeeb109659750ad8153ce | 288ef8ff952e44eb86ae1471437543e8aa29651d | Fix intel mac build
Make sure we're building an x86 ext_server lib when cross-compiling | [
{
"path": "llm/dyn_ext_server.go",
"patch": "@@ -75,7 +75,7 @@ func newDynExtServer(library, model string, adapters, projectors []string, opts\n \tupdatePath(filepath.Dir(library))\n \tlibPath := C.CString(library)\n \tdefer C.free(unsafe.Pointer(libPath))\n-\tresp := newExtServerResp(128)\n+\tresp := newEx... | 2024-01-13T22:46:34 |
ggml-org/llama.cpp | dbb852b549adf29609ec53b518f7922a982f14b9 | 5f55c385cbff7bcee0901782dec6d01de53f4dfe | ggml-cpu: arm64: q4_K repack gemm and gemv implementations (i8mm) (#16739)
* Enabled q4_K_8x8_q8_K path on ARM
* wip: I8mm qs multiplication, pending bias
* cpu : arm : REPACK gemm q4_K8x8 implementation
Signed-off-by: Alberto Cabrera <alberto.cabrera@liquid.ai>
* Guard gemm with proper features, improved superblo... | [
{
"path": "ggml/src/ggml-cpu/arch-fallback.h",
"patch": "@@ -51,10 +51,8 @@\n #elif defined(__aarch64__) || defined(__arm__) || defined(_M_ARM) || defined(_M_ARM64)\n // repack.cpp\n #define ggml_quantize_mat_q8_K_4x8_generic ggml_quantize_mat_q8_K_4x8\n-#define ggml_gemv_q4_K_8x8_q8_K_generic ggml_gemv_q4_... | 2025-11-24T11:08:11 |
vuejs/vue | d4d553ced75d8c73e75b85cec398be4b09f6f669 | 41d774d112946f986bf0b0e3f30fd962c01ceba2 | fix(v-model): avoid unnecessary change event on select options change
fix #6193
via #6194 | [
{
"path": "src/platforms/web/runtime/directives/model.js",
"patch": "@@ -53,7 +53,14 @@ export default {\n const prevOptions = el._vOptions\n const curOptions = el._vOptions = [].map.call(el.options, getValue)\n if (curOptions.some((o, i) => !looseEqual(o, prevOptions[i]))) {\n- tri... | 2017-08-29T16:52:19 |
ggml-org/llama.cpp | 5f55c385cbff7bcee0901782dec6d01de53f4dfe | 4902eebe33ed2341ef9bd7c80195bbd9d24f4d5f | ggml: add RISC-V cpu-feats (#17461)
* ggml: add RISC-V cpu-feats
Signed-off-by: Wang Yang <yangwang@iscas.ac.cn>
* fix comment[1]
---------
Signed-off-by: Wang Yang <yangwang@iscas.ac.cn> | [
{
"path": "ggml/src/CMakeLists.txt",
"patch": "@@ -328,6 +328,14 @@ function(ggml_add_cpu_backend_variant tag_name)\n set(GGML_INTERNAL_${feat} OFF)\n endforeach()\n \n+ foreach (feat ${ARGN})\n+ set(GGML_INTERNAL_${feat} ON)\n+ endforeach()\n+ elseif (GGML_SY... | 2025-11-24T11:07:14 |
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.