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 | b97521614ad47b04fb4b5286582efcd984ced0b7 | 534cbf8a5d332ad8cd26124c5a0a9201f73ca6d7 | Fix RT-DETR cache for generate_anchors (#31671)
* Fix cache and type conversion
* Add test
* Fixup
* nit
* [run slow] rt_detr
* Fix test
* Fixup
* [run slow] rt_detr
* Update src/transformers/models/rt_detr/modeling_rt_detr.py | [
{
"path": "src/transformers/models/rt_detr/modeling_rt_detr.py",
"patch": "@@ -1656,7 +1656,11 @@ def unfreeze_backbone(self):\n param.requires_grad_(True)\n \n @lru_cache(maxsize=32)\n- def generate_anchors(self, spatial_shapes=None, grid_size=0.05, dtype=torch.float32, device=\"cpu\"):\... | 2024-07-03T13:19:57 |
rust-lang/rust | 224dad154bd055bf3fb792ae51f9c6d91a927f9b | 90ebc246075f9b56cd0a2dee2496698a94c9c033 | Fix formatting. | [
{
"path": "src/tools/unicode-table-generator/src/skiplist.rs",
"patch": "@@ -68,10 +68,7 @@ impl RawEmitter {\n assert!(inserted);\n }\n \n- writeln!(\n- &mut self.file,\n- \"use super::ShortOffsetRunHeader;\\n\"\n- ).unwrap();\n+ writeln!(&mut ... | 2025-03-08T11:47:40 |
golang/go | 3089ef6bd7ecc7af4b23eb68e3d7879d340aa673 | 15f01d6ae9853fd51ee8842d9af93d04ce25458c | cmd/asm: add several arm64 SIMD instructions
This patch enables VSLI, VUADDW(2), VUSRA and FMOVQ SIMD instructions
required by the issue #40725. And the GNU syntax of 'FMOVQ' is 128-bit
ldr/str(immediate, simd&fp).
Add test cases.
Fixes #40725
Change-Id: Ide968ef4a9385ce4cd8f69bce854289014d30456
Reviewed-on: https:... | [
{
"path": "src/cmd/asm/internal/asm/testdata/arm64.s",
"patch": "@@ -454,6 +454,17 @@ TEXT\tfoo(SB), DUPOK|NOSPLIT, $-8\n \tFMOVD\tF20, (R2) // 540000fd\n \tFMOVD.P\tF20, 8(R1) // 348400fc\n \tFMOVD.W\t8(R1), F20 ... | 2020-09-28T10:58:51 |
ollama/ollama | ba26c7aa00635c78c2028aba680556801c02fd64 | 63c763685f1dc94f7efe4742b00b226be99505d0 | it will always return an error due to Kill() discarding Wait() errors | [
{
"path": "llm/server.go",
"patch": "@@ -903,9 +903,7 @@ func (s *llmServer) Close() error {\n \t\t\treturn err\n \t\t}\n \n-\t\tif err := s.cmd.Wait(); err != nil {\n-\t\t\treturn err\n-\t\t}\n+\t\t_ = s.cmd.Wait()\n \n \t\tslog.Debug(\"llama server stopped\")\n \t}",
"additions": 1,
"deletions": 3... | 2024-04-29T22:48:14 |
ggml-org/llama.cpp | ab9d4c3678a6f8bb797610a27bc0af493fcf786c | 1af9dab32b555f7e3aac8b1bfe4e4cb780cb4185 | server : improve mtmd ctx checkpoints (#20726)
* server : improve mtmd ctx checkpoints
* server : fix off-by-one in pos_min_thold | [
{
"path": "tools/server/server-context.cpp",
"patch": "@@ -2307,8 +2307,8 @@ struct server_context_impl {\n \n llama_pos pos_next = slot.prompt.tokens.pos_next(n_past);\n \n- // note: when n_swa == 0, the model does not use SWA, which is equivalent to a... | 2026-03-20T09:13:12 |
huggingface/transformers | 534cbf8a5d332ad8cd26124c5a0a9201f73ca6d7 | 65a02cd27dd3cd73e5662a50b646d7baaaab8fad | [fix bug] logits's shape different from label's shape in preprocess_logits_for_metrics (#31447)
* [fix BUG] pad labels before use it in preprocess_logits_for_metrics
* a more readable fix
labels can't use `gather` before pass to `preprocess_logits_for_metrics`, so must split into 2 if-block
* add a comment
... | [
{
"path": "src/transformers/trainer.py",
"patch": "@@ -3839,6 +3839,9 @@ def evaluation_loop(\n inputs_decode = self.gather_function((inputs_decode))\n if not self.args.batch_eval_metrics or description == \"Prediction\":\n all_inputs.add(inputs_decode)\n+... | 2024-07-03T10:58:27 |
nodejs/node | bd5acfe44d8344306a91c580292c6106f6a98d7d | a4a9a3de6ecceeff915fb79aa43457b3b3a3fe59 | test: use common.crashOnUnhandledRejection
Add common.crashOnUnhandledRejection to
test-microtask-queue-run-immediate.js
PR-URL: https://github.com/nodejs/node/pull/17214
Reviewed-By: Myles Borins <myles.borins@gmail.com>
Reviewed-By: Daniel Bevenius <daniel.bevenius@gmail.com>
Reviewed-By: Gireesh Punathil <gpunathi... | [
{
"path": "test/parallel/test-microtask-queue-run-immediate.js",
"patch": "@@ -20,9 +20,11 @@\n // USE OR OTHER DEALINGS IN THE SOFTWARE.\n \n 'use strict';\n-require('../common');\n+const common = require('../common');\n const assert = require('assert');\n \n+common.crashOnUnhandledRejection();\n+\n functi... | 2017-11-22T08:34:04 |
huggingface/transformers | 65a02cd27dd3cd73e5662a50b646d7baaaab8fad | ddfaf1192629f7efadeb8caea93c77ffd98573cb | Add ignore_errors=True to trainer.py rmtree in _inner_training_loop (#31668)
Update trainer.py | [
{
"path": "src/transformers/trainer.py",
"patch": "@@ -2422,7 +2422,7 @@ def _inner_training_loop(\n for checkpoint in checkpoints_sorted:\n if not os.path.samefile(checkpoint, self.state.best_model_checkpoint):\n logger.info(f\"Deleting older checkpoint [{che... | 2024-07-03T10:54:49 |
ollama/ollama | 63c763685f1dc94f7efe4742b00b226be99505d0 | 34a4a94f13d283ef794ba02f84ded96a794bb5d2 | log when the waiting for the process to stop to help debug when other tasks execute during this wait.
expire timer clear the timer reference because it will not be reused.
close will clean up expireTimer if calling code has not already done this. | [
{
"path": "llm/server.go",
"patch": "@@ -902,7 +902,12 @@ func (s *llmServer) Close() error {\n \t\tif err := s.cmd.Process.Kill(); err != nil {\n \t\t\treturn err\n \t\t}\n-\t\treturn s.cmd.Wait()\n+\n+\t\tif err := s.cmd.Wait(); err != nil {\n+\t\t\treturn err\n+\t\t}\n+\n+\t\tslog.Debug(\"llama server st... | 2024-04-29T18:06:56 |
ggml-org/llama.cpp | 6d99b44c7eb868e8b9d49d0b4cab243dd70a4b53 | 464fd0e71f13b9ed7676965faa35f7aee02d8c86 | docs : fix Metal backend op support status in ops.md (#20779)
Regenerate docs/ops/Metal.csv using test-backend-ops on Apple M5
and rebuild docs/ops.md via scripts/create_ops_docs.py.
Five ops were incorrectly marked as not supported (❌) for Metal:
- DIAG: ❌ → ✅
- POOL_1D: ❌ → ✅
- SET: ❌ → ... | [
{
"path": "docs/ops.md",
"patch": "@@ -12,9 +12,9 @@ Legend:\n - 🟡 Partially supported by this backend\n - ❌ Not supported by this backend\n \n-| Operation | BLAS | CANN | CPU | CUDA | Metal | OpenCL | SYCL | Vulkan | WebGPU | ZenDNN | zDNN |\n+| Operation | BLAS | CANN | CPU | CUDA | MTL | OpenCL | SYCL |... | 2026-03-20T09:06:38 |
rust-lang/rust | ff674232d198b196c8fa9d2092b8165b4bcce6b2 | 871be8f54ace198c42780b366d1817ffc77e9fa6 | Fix checkout in post-merge workflow | [
{
"path": ".github/workflows/post-merge.yml",
"patch": "@@ -16,6 +16,11 @@ jobs:\n pull-requests: write\n steps:\n - uses: actions/checkout@v4\n+ with:\n+ # Make sure that we have enough commits to find the parent merge commit.\n+ # Since all merges should be through... | 2025-03-08T11:46:33 |
huggingface/transformers | 7f91f168a14c7a305252675730b3248a1c987944 | f91c16d270e5e3ff32fdb32ccf286d05c03dfa66 | fix assisted decoding (#31401)
* fix assisted decoding
* check None
* fix typo
* fix _prepare_special_tokens
* fix style
* fix lint
* add tests for assisted decoding
* fix style
* fix tests check | [
{
"path": "src/transformers/generation/utils.py",
"patch": "@@ -1493,8 +1493,11 @@ def _tensor_or_none(token_kwargs, token_self, device=None):\n device = self.device\n \n token = token_kwargs if token_kwargs is not None else token_self\n- if token is None or isinstance... | 2024-07-03T08:22:56 |
ggml-org/llama.cpp | 21c804521479c750bdf6d26506964ab3aa207481 | c46583b86bed573c4ff30685dae59874f124e664 | jinja : fix heap OOB read in value equality comparison (#20782)
Address GHSA-q9j6-4hhc-rq9p and GHSA-2q4c-9gq5-5vfp.
The three-iterator overload of std::equal in value_array_t::equivalent()
and value_object_t::equivalent() reads past the end of the shorter
container when comparing arrays or objects of different lengt... | [
{
"path": "common/jinja/value.h",
"patch": "@@ -451,7 +451,7 @@ struct value_array_t : public value_t {\n }\n protected:\n virtual bool equivalent(const value_t & other) const override {\n- return typeid(*this) == typeid(other) && is_hashable() && other.is_hashable() && std::equal(val_arr.beg... | 2026-03-20T06:15:17 |
ollama/ollama | 34a4a94f13d283ef794ba02f84ded96a794bb5d2 | f4a73d57a4e3182a4dd25dae93a779a13423dfc0 | ignore debug bin files | [
{
"path": ".gitignore",
"patch": "@@ -11,4 +11,5 @@ ggml-metal.metal\n .idea\n test_data\n *.crt\n-llm/build\n\\ No newline at end of file\n+llm/build\n+__debug_bin*\n\\ No newline at end of file",
"additions": 2,
"deletions": 1,
"language": "Unknown"
}
] | 2024-04-29T15:05:54 |
rust-lang/rust | b8a217081d0d7aef0c89c9dacc21c020faa0949b | f79f3d31a3cbf87e50531851f5079e7fcba7ce81 | Refactor coroutine layout logic to precompute all sublayouts
Also properly attaches spans on layouts of non-promoted coroutine
locals, which slightly improves the error messages for some coroutine tests. | [
{
"path": "compiler/rustc_ty_utils/src/layout.rs",
"patch": "@@ -5,23 +5,21 @@ use hir::def_id::DefId;\n use rustc_abi::Integer::{I8, I32};\n use rustc_abi::Primitive::{self, Float, Int, Pointer};\n use rustc_abi::{\n- AddressSpace, BackendRepr, FIRST_VARIANT, FieldIdx, FieldsShape, HasDataLayout, Layout... | 2025-03-07T21:09:56 |
huggingface/transformers | f91c16d270e5e3ff32fdb32ccf286d05c03dfa66 | cd0935dd55dcc1ba8d20305378d5f4e11126d225 | Fix documentation for Gemma2. (#31682)
* Fix documentation for Gemma2.
Model sizes and Blog post URL are wrong in the documentation.
* Update docs/source/en/model_doc/gemma2.md
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>
---------
Co-authored-by: amyeroberts <22614925+am... | [
{
"path": "docs/source/en/model_doc/gemma2.md",
"patch": "@@ -19,12 +19,12 @@ rendered properly in your Markdown viewer.\n \n ## Overview\n \n-The Gemma2 model was proposed in [Gemma2: Open Models Based on Gemini Technology and Research](https://blog.google/technology/developers/Gemma2-open-models/) by Gemm... | 2024-07-02T22:04:53 |
ggml-org/llama.cpp | c46583b86bed573c4ff30685dae59874f124e664 | c1b911654a581662c19eb51829705e4ed7020673 | common/parser : fix out_of_range crash in throw path (#20424 regression) (#20777)
* chat : fix out_of_range crash in throw path (#20424 regression)
#20424 introduced effective_input = generation_prompt + input, but the
throw path uses input.substr(result.end) where result.end is a position
within effective_input. Eve... | [
{
"path": "common/chat.cpp",
"patch": "@@ -1774,7 +1774,7 @@ common_chat_msg common_chat_peg_parse(const common_peg_arena & src_pars\n return msg;\n }\n throw std::runtime_error(std::string(\"Failed to parse input at pos \") + std::to_string(result.end) + \": \" +\n- ... | 2026-03-20T01:37:22 |
ollama/ollama | f4a73d57a4e3182a4dd25dae93a779a13423dfc0 | 948114e3e3325af01c31697dee861459624ca805 | fix runner expire during active use. Clearing the expire timer as it is used. Allowing the finish to assign an expire timer so that the runner will expire after no use. | [
{
"path": "server/sched.go",
"patch": "@@ -296,6 +296,10 @@ func (pending *LlmRequest) useLoadedRunner(runner *runnerRef, finished chan *Llm\n \trunner.refMu.Lock()\n \tdefer runner.refMu.Unlock()\n \trunner.refCount++\n+\tif runner.expireTimer != nil {\n+\t\trunner.expireTimer.Stop()\n+\t\trunner.expireTim... | 2024-04-28T16:43:40 |
golang/go | 15f01d6ae9853fd51ee8842d9af93d04ce25458c | 7fe2a84834537b58578791dd041b7bb40572620a | cmd/compile: delay expansion of OpArg until expand_calls
As it says, delay expanpsion of OpArg to the expand_calls phase,
to enable (eventually) interprocedural SSA optimizations, and
(sooner) change to a register ABI.
Includes a round of cleanup to function names and comments,
largely to match the expanded scope of ... | [
{
"path": "src/cmd/compile/fmtmap_test.go",
"patch": "@@ -136,7 +136,6 @@ var knownFormats = map[string]string{\n \t\"cmd/compile/internal/types.EType %s\": \"\",\n \t\"cmd/compile/internal/types.EType %v\": \"\",\n \t\"cmd/internal/obj.ABI %v\": \"\",\n-\t\"c... | 2020-10-13T23:24:04 |
rust-lang/rust | f83af2aa6b395c2e6c5639e8536f08f24d3f2128 | 20f0108ada90e37c2350ad68fdc6afe56f51539d | bootstrap: Fix stack printing when a step cycle is detected | [
{
"path": "src/bootstrap/src/core/build_steps/run.rs",
"patch": "@@ -367,3 +367,28 @@ impl Step for FeaturesStatusDump {\n cmd.run(builder);\n }\n }\n+\n+/// Dummy step that can be used to deliberately trigger bootstrap's step cycle\n+/// detector, for automated and manual testing.\n+#[derive(Cl... | 2025-03-08T07:11:47 |
huggingface/transformers | 82486e5995ed0a65520b10ce1ea938214a199231 | a9701953ffefcf8f10ec5a94dc2c55633648ba28 | 🚨🚨 TextGenerationPipeline: rely on the tokenizer default kwargs (#31747)
* rely on the tokenizer default kwargs
* fix a few tests | [
{
"path": "src/transformers/pipelines/text_generation.py",
"patch": "@@ -266,31 +266,33 @@ def preprocess(\n prompt_text,\n prefix=\"\",\n handle_long_generation=None,\n- add_special_tokens=False,\n+ add_special_tokens=None,\n truncation=None,\n- padding=... | 2024-07-02T14:17:42 |
ggml-org/llama.cpp | c1b911654a581662c19eb51829705e4ed7020673 | b739738dadf0b66a59546d7240c554d61c07c2f0 | server: fix router mode deadlock on child crash and TOCTOU race in models_max (#20763)
Two bugs in `server_models::load()` that affect router mode reliability:
**Bug 1: Deadlock when child process crashes**
When a child process is killed (e.g., SIGKILL from OS code signature
validation), the monitoring thread deadlo... | [
{
"path": "tools/server/server-models.cpp",
"patch": "@@ -539,6 +539,22 @@ void server_models::load(const std::string & name) {\n return;\n }\n \n+ // Re-check capacity under the lock to prevent concurrent loads from\n+ // exceeding models_max. Without this, the window between unload_lru()... | 2026-03-19T21:16:05 |
nodejs/node | 446c1ecfdaf99de38082d048626f301109da49c6 | 2346c6389f2253b9fc4f29c0e552283ab305f702 | src, tools: add debug symbols for node internals
Before these changes, only V8 added debug symbols to Node's binary,
limiting the possibilities for debugger's developers to add some
features that rely on investigating Node's internal structures.
These changes are a first steps towards empowering debug tools to
naviga... | [
{
"path": "node.gyp",
"patch": "@@ -178,6 +178,7 @@\n \n 'dependencies': [\n 'node_js2c#host',\n+ 'deps/nghttp2/nghttp2.gyp:nghttp2'\n ],\n \n 'includes': [\n@@ -298,6 +299,7 @@\n # node.gyp is added to the project by default.\n 'common.gypi',\n '<(SH... | 2017-09-24T22:34:19 |
ollama/ollama | 948114e3e3325af01c31697dee861459624ca805 | a3e60d9058d4459aac3a06f7cab80b0e894e1b38 | fix sched to wait for the runner to terminate to ensure following vram check will be more accurate | [
{
"path": "llm/server.go",
"patch": "@@ -899,7 +899,10 @@ func (s *llmServer) Detokenize(ctx context.Context, tokens []int) (string, error\n func (s *llmServer) Close() error {\n \tif s.cmd != nil {\n \t\tslog.Debug(\"stopping llama server\")\n-\t\treturn s.cmd.Process.Kill()\n+\t\tif err := s.cmd.Process.K... | 2024-04-28T16:41:38 |
golang/go | 7fe2a84834537b58578791dd041b7bb40572620a | 308ec220a2e55e0948505881b051cbefa05cf696 | strconv: remove extfloat.go atof code path
Prior to this commit, strconv.ParseFloat (known in C as atof) takes the
first of four algorithms to succeed: atof64exact, eiselLemire64,
extFloat, fallback. The Eisel-Lemire implementation is a recent addition
but, now that it exists, the extFloat implementation (based on the... | [
{
"path": "src/strconv/atof.go",
"patch": "@@ -576,24 +576,14 @@ func atof32(s string) (f float32, n int, err error) {\n \t\treturn float32(f), n, err\n \t}\n \n-\tif optimize {\n-\t\t// Try pure floating-point arithmetic conversion.\n-\t\tif !trunc {\n-\t\t\tif f, ok := atof32exact(mantissa, exp, neg); ok ... | 2020-10-23T01:11:35 |
ggml-org/llama.cpp | 340807273b6aa765c9353804b7ce680920373cb6 | 26c9ce128825ba53a91baf75b5b817a1373b46bf | hip: Avoid compiler bug in RDNA code generation during debug builds on Windows (#20655) | [
{
"path": "ggml/src/ggml-hip/CMakeLists.txt",
"patch": "@@ -129,6 +129,11 @@ endif()\n \n if (CXX_IS_HIPCC)\n set_source_files_properties(${GGML_SOURCES_ROCM} PROPERTIES LANGUAGE CXX)\n+ if (WIN32 AND CMAKE_BUILD_TYPE STREQUAL \"Debug\")\n+ # CMake on Windows doesn't support the HIP language y... | 2026-03-19T18:14:08 |
nodejs/node | bb31a1689fc5aa0c26372b9096e4e4378e40f205 | 6a87cac55084ef96c148986305acd26303d61f13 | src: fix compiler warning
This commit fixes a -Winconsistent-missing-override warning.
PR-URL: https://github.com/nodejs/node/pull/17195
Reviewed-By: Refael Ackermann <refack@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com> | [
{
"path": "src/node_http2.h",
"patch": "@@ -555,7 +555,7 @@ class Http2Stream : public AsyncWrap,\n int ReadStop() override;\n \n // Required for StreamBase\n- int DoShutdown(ShutdownWrap* req_wrap);\n+ int DoShutdown(ShutdownWrap* req_wrap) override;\n \n // Initiate a response on this stream.\n ... | 2017-11-21T15:34:59 |
huggingface/transformers | a9701953ffefcf8f10ec5a94dc2c55633648ba28 | 57d7594a79a9f5d835abf2d4d384db0e4818e548 | [whisper] static kv cache (#31166)
* make work with cache abstraction
* correct for static cache
* hacks for compile
* make fast
* fix
* fix pos ids
* generate
* fix sdpa
* fix sdpa cache pos
* fix fa2
* clean fa2
* integrate cache into generate
* make style
* copies
* more co... | [
{
"path": "docs/source/en/internal/generation_utils.md",
"patch": "@@ -391,6 +391,12 @@ A [`Constraint`] can be used to force the generation to include specific tokens\n - get_seq_length\n - reset\n \n+[[autodoc]] EncoderDecoderCache\n+ - get_seq_length\n+ - to_legacy_cache\n+ - from_legacy... | 2024-07-02T12:24:15 |
ollama/ollama | a3e60d9058d4459aac3a06f7cab80b0e894e1b38 | f0c454ab577f4cf716e23ff56cbc3c316c88a78c | README.md: fix typos (#4007)
Co-authored-by: Blake Mizerany <blake.mizerany@gmail.com> | [
{
"path": "README.md",
"patch": "@@ -1,5 +1,5 @@\n <div align=\"center\">\n- <img alt=\"ollama\" height=\"200px\" src=\"https://github.com/ollama/ollama/assets/3325447/0d0b44e2-8f4a-4e99-9b52-a5c1c741c8f7\">\n+ <img alt=\"ollama\" height=\"200px\" src=\"https://github.com/ollama/ollama/assets/3325447/0d0b4... | 2024-05-01T17:39:38 |
ggml-org/llama.cpp | 26c9ce128825ba53a91baf75b5b817a1373b46bf | 76f2dc70c360d6506c588d68b58ff14d0120ce8b | server: Add cached_tokens info to oaicompat responses (#19361)
* tests : fix fetch_server_test_models.py
* server: to_json_oaicompat cached_tokens
Adds OpenAI and Anthropic compatible information about the
number of cached prompt tokens used in a response. | [
{
"path": "scripts/fetch_server_test_models.py",
"patch": "@@ -95,9 +95,9 @@ def collect_hf_model_test_parameters(test_file) -> Generator[HuggingFaceModel, N\n '-p', 'Hey',\n '--no-warmup',\n '--log-disable',\n- '-no-cnv']\n+ '-st']\n if m.hf... | 2026-03-19T18:09:33 |
golang/go | d8044a6744f9570e964c6e35e12827473ff12cde | c1afbf69c71bc624a4766a48ef637a5f726dfe4e | database/sql: fix TestTxStmtDeadlock test
Drop error check because errors can be
not only ErrTxDone for tx stmt executions,
and the purpose of the test is just reproducing
deadlock.
Fixes #42259
Change-Id: I9e7105ada1403ec7064dcc1c3385b36893a1c195
Reviewed-on: https://go-review.googlesource.com/c/go/+/266097
Reviewe... | [
{
"path": "src/database/sql/sql_test.go",
"patch": "@@ -2828,12 +2828,10 @@ func TestTxStmtDeadlock(t *testing.T) {\n \t}\n \t// Run number of stmt queries to reproduce deadlock from context cancel\n \tfor i := 0; i < 1e3; i++ {\n+\t\t// Encounter any close related errors (e.g. ErrTxDone, stmt is closed)\n+... | 2020-10-28T22:03:08 |
huggingface/transformers | 57d7594a79a9f5d835abf2d4d384db0e4818e548 | 93cd94b79d62a8a8b6f4f20e51fda95e6daa4d3a | Fix mistral ONNX export (#31696)
* use bitwise or
* why is the CI not triggered? | [
{
"path": "src/transformers/models/mistral/modeling_mistral.py",
"patch": "@@ -1089,8 +1089,9 @@ def _update_causal_mask(\n exclude_mask = torch.arange(target_length, device=device) > cache_position.reshape(-1, 1)\n if self.config.sliding_window is not None:\n if not ... | 2024-07-02T11:54:10 |
nodejs/node | 6a87cac55084ef96c148986305acd26303d61f13 | 8578e81b2b41eac6aea7a812b8e5f35a7362a6df | test: remove unused parameter in test-next-tick-error-spin.js
PR-URL: https://github.com/nodejs/node/pull/17185
Reviewed-By: Anatoli Papirovski <apapirovski@mac.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Michaël Zasso <targos@protonmail.com>
Reviewed-By: Gireesh Punathil <gpunathi@in.ibm.com>
Revie... | [
{
"path": "test/sequential/test-next-tick-error-spin.js",
"patch": "@@ -42,7 +42,7 @@ if (process.argv[2] !== 'child') {\n process.maxTickDepth = 10;\n \n // in the error handler, we trigger several MakeCallback events\n- d.on('error', function(e) {\n+ d.on('error', function() {\n console.log('a')... | 2017-11-21T12:09:12 |
ollama/ollama | bd8eed57fc7d5f9b9b9d333b9c395864fb2378d8 | 9cf0f2e9736c31bbd00f4671613d0e31ecb3c4ea | fix parser name | [
{
"path": "parser/parser.go",
"patch": "@@ -27,8 +27,9 @@ const (\n )\n \n var (\n-\terrMissingFrom = errors.New(\"no FROM line\")\n-\terrInvalidRole = errors.New(\"role must be one of \\\"system\\\", \\\"user\\\", or \\\"assistant\\\"\")\n+\terrMissingFrom = errors.New(\"no FROM line\")\n+\terrInval... | 2024-04-27T00:11:47 |
ggml-org/llama.cpp | 74c42ee1f4f0fa3609c8aef543edb6f307826063 | b49d8b875724ffa94d09c4abf933decae0c33f07 | hexagon: add Matrix Extensions (HMX) for Hexagon NPU backend (#20693)
* migrate(vtcm): unify VTCM management for HMX merge
- Add HMX fields to htp_context (#ifdef HTP_HAS_HMX): hmx_enabled,
hmx_dma, vtcm_scratch_size, exp2_table
- Add HTP_VTCM_SESSION_HOLD CMake option (default ON): hold VTCM for
entire session i... | [
{
"path": "ggml/src/ggml-hexagon/ggml-hexagon.cpp",
"patch": "@@ -45,6 +45,7 @@ static int opt_verbose = 0;\n static int opt_profile = 0;\n static int opt_hostbuf = 1; // hostbuf ON by default\n static int opt_experimental = 0;\n+static int opt_use_hmx = 1; // when set, en... | 2026-03-19T16:11:06 |
golang/go | 07d206f7698f4d45544a2f9b6051ede594ba04cc | 87d59bcdc354b035e17a1dadbfc4883bbcac5bf8 | cmd/go: use internal/testenv instead of computing canRun and skipExternal ad-hoc
Fixes #42223
Change-Id: Icf9bb61d48f0a6c7fd6f74e80e333a4837aa52ab
Reviewed-on: https://go-review.googlesource.com/c/go/+/265781
Trust: Bryan C. Mills <bcmills@google.com>
Run-TryBot: Bryan C. Mills <bcmills@google.com>
Reviewed-by: Ian L... | [
{
"path": "src/cmd/go/go_test.go",
"patch": "@@ -35,72 +35,29 @@ import (\n )\n \n var (\n-\tcanRun = true // whether we can run go or ./testgo\n \tcanRace = false // whether we can run the race detector\n \tcanCgo = false // whether we can use cgo\n \tcanMSan = false // whether we can run the memory san... | 2020-10-28T14:44:24 |
huggingface/transformers | cf85e86e9ad22ddd02c519d21c4783e3dbf72178 | 3345ae733b6f4aeb7204a0f3e646a3cdbaad0023 | remove incorrect urls pointing to the llava repository (#31107)
* remove incorrect urls pointing to the llava repository
* remove incorrect urls pointing to the llava repository; removing entire comments
* remove incorrect urls pointing to the llava repository; removing entire comments; ran fix-copies
* ran f... | [
{
"path": "src/transformers/models/idefics2/modeling_idefics2.py",
"patch": "@@ -1344,9 +1344,6 @@ class Idefics2PreTrainedModel(PreTrainedModel):\n _supports_cache_class = True\n \n def _init_weights(self, module):\n- # important: this ported version of Idefics2 isn't meant for training from... | 2024-07-02T11:24:55 |
nodejs/node | f005656719326bd68e12cda0c0af78756ebd11c6 | 2540e3dacfe7ae55f4b3f2c229d8486d88a0622c | src: implement v8::TaskRunner API in NodePlatform
V8 is switching APIs for scheduling tasks. Implement the new APIs.
Fixes: https://github.com/nodejs/node-v8/issues/24
Refs: https://github.com/v8/v8/commit/c690f54d9580243c53f7d892fcff1ce6bae4bfc0
PR-URL: https://github.com/nodejs/node/pull/17134
Fixes: https://githu... | [
{
"path": "src/node_platform.cc",
"patch": "@@ -24,6 +24,43 @@ static void BackgroundRunner(void* data) {\n }\n }\n \n+BackgroundTaskRunner::BackgroundTaskRunner(int thread_pool_size) {\n+ for (int i = 0; i < thread_pool_size; i++) {\n+ std::unique_ptr<uv_thread_t> t { new uv_thread_t() };\n+ if (u... | 2017-11-19T13:29:08 |
ollama/ollama | 8907bf51d235e86854c762870f203380866f1ae3 | abe614c705736eed06440d657ab75ca094fc78f3 | fix multiline | [
{
"path": "parser/parser.go",
"patch": "@@ -110,11 +110,16 @@ func Parse(r io.Reader) (cmds []Command, err error) {\n \tcase stateComment, stateNil:\n \t\t// pass; nothing to flush\n \tcase stateValue:\n-\t\tif _, ok := unquote(b.String()); !ok {\n+\t\ts, ok := unquote(b.String())\n+\t\tif !ok {\n \t\t\tret... | 2024-04-25T02:17:26 |
golang/go | 642329fdd55aabafc67b3a7c50902e29125621ab | e3c58bbeb8c76fa3abc0f7153edbab72208c1f88 | Revert "cmd/compile: split exported/non-exported methods for interface type"
This reverts commit 8f26b57f9afc238bdecb9b7030bc2f4364093885.
Reason for revert: break a bunch of code, include standard library.
Fixes #42123
Change-Id: Ife90ecbafd2cb395623d1db555fbfc9c1b0098e0
Reviewed-on: https://go-review.googlesource... | [
{
"path": "doc/go1.16.html",
"patch": "@@ -264,15 +264,6 @@ <h3 id=\"net\"><a href=\"/pkg/net/\">net</a></h3>\n On Linux kernel version 4.1 and above, the maximum is now <code>4294967295</code>.\n </p>\n \n-<h3 id=\"reflect\"><a href=\"/pkg/reflect/\">reflect</a></h3>\n-\n-<p><!-- CL 259237, golang.org/is... | 2020-10-21T17:25:17 |
ggml-org/llama.cpp | 5e54d51b199ad2d70cf6eba4bff756bbf63366a6 | c1258830b28dddc65ac3aa3452214553fdece30c | common/parser: add proper reasoning tag prefill reading (#20424)
* Implement proper prefill extraction
* Refactor cli parameters, update docs, move reasoning budget sampler part to common/reasoning-budget.cpp
* Update tools/server/server-task.cpp
* refactor: move grammars to variant, remove grammar_external, handle... | [
{
"path": "common/arg.cpp",
"patch": "@@ -1830,23 +1830,23 @@ common_params_context common_params_parser_init(common_params & params, llama_ex\n ).set_sparam());\n add_opt(common_arg(\n {\"--grammar\"}, \"GRAMMAR\",\n- string_format(\"BNF-like grammar to constrain generations (see sam... | 2026-03-19T15:58:21 |
huggingface/transformers | cb298978ade3f1edb0ffd02ee079a69f08917a2a | 82a1fc7256bf27f83aec3a93543b6d156add09cf | add gather_use_object arguments (#31514)
* add gather_use_object arguments
* fix name and pass the CI test for Seq2SeqTrainer
* make style
* make it to functools
* fix typo
* add accelerate version:
* adding warning
* Update src/transformers/trainer.py
Co-authored-by: Marc Sun <57196510+SunMarc... | [
{
"path": "src/transformers/trainer.py",
"patch": "@@ -4605,6 +4605,11 @@ def create_accelerator_and_postprocess(self):\n # some Trainer classes need to use `gather` instead of `gather_for_metrics`, thus we store a flag\n self.gather_function = self.accelerator.gather_for_metrics\n \n+ ... | 2024-06-28T12:50:27 |
nodejs/node | 2540e3dacfe7ae55f4b3f2c229d8486d88a0622c | 5d35a1a3389c0e93be44a7458bc35a2a3cc9fb02 | deps: cherry-pick 98c40a4bae915 from V8 upstream
Original commit message:
[platform] Return task runners as shared_ptr
At the moment, task runners are returned as unique_ptr. This is
inconvenient, however. In all implementations I did, the platform holds
a shared pointer of the task runner and wraps ... | [
{
"path": "deps/v8/include/v8-platform.h",
"patch": "@@ -199,7 +199,7 @@ class Platform {\n * Returns a TaskRunner which can be used to post a task on the foreground.\n * This function should only be called from a foreground thread.\n */\n- virtual std::unique_ptr<v8::TaskRunner> GetForegroundTask... | 2017-11-19T14:02:51 |
ollama/ollama | 0a7fdbe533b087c47079a2cc54de6bf916443cbc | 5950c176caf585b04ac517003d1385833e5e4927 | prompt to display and add local ollama keys to account (#3717)
- return descriptive error messages when unauthorized to create blob or push a model
- display the local public key associated with the request that was denied | [
{
"path": "auth/auth.go",
"patch": "@@ -10,12 +10,44 @@ import (\n \t\"log/slog\"\n \t\"os\"\n \t\"path/filepath\"\n+\t\"strings\"\n \n \t\"golang.org/x/crypto/ssh\"\n )\n \n const defaultPrivateKey = \"id_ed25519\"\n \n+func keyPath() (string, error) {\n+\thome, err := os.UserHomeDir()\n+\tif err != nil {\... | 2024-04-30T18:02:08 |
ggml-org/llama.cpp | c1258830b28dddc65ac3aa3452214553fdece30c | 922b90e5672c869614baf6c617ed1069bab9a8e8 | ggml webgpu: ops support for qwen3.5 (SET, TRI_SOLVE, SSM_CONV, GATED_DELTA_NET) + GET_ROWS optimization (#20687)
* Implement l2_norm, set, tri
* Add DIAG/SOLVE_TRI
* Add SSM_CONV
* Better get_rows and gated_delta_net to support qwen3.5
* Clean up, update ops.md
* Fix binding_index type for wasm
* Fix read write... | [
{
"path": "docs/ops.md",
"patch": "@@ -47,7 +47,7 @@ Legend:\n | FILL | ❌ | ❌ | ✅ | ✅ | ✅ | ❌ | ❌ | ✅ | ✅ | ❌ | ❌ |\n | FLASH_ATTN_EXT | ❌ | 🟡 | ✅ | 🟡 | 🟡 | 🟡 | 🟡 | 🟡 | 🟡 | ❌ | ❌ |\n | FLOOR | ❌ | ❌ | ✅ | 🟡 | ❌ | ❌ | 🟡 | 🟡 | ... | 2026-03-19T15:45:28 |
golang/go | e3c58bbeb8c76fa3abc0f7153edbab72208c1f88 | d4c1ad882973e407ff85b977f4ce5b9435451190 | os: do not use procfs for os.Executable in dragonfly
procfs(5) is not always mounted in DragonFly BSD, for example during
the binary package build with synth. os.Executable() consumers
will then fail, we've spotted this when trying to build tinygo:
[...]
copying source files
./build/tinygo build-bu... | [
{
"path": "src/os/executable_dragonfly.go",
"patch": "@@ -0,0 +1,12 @@\n+// Copyright 2020 The Go Authors. All rights reserved.\n+// Use of this source code is governed by a BSD-style\n+// license that can be found in the LICENSE file.\n+\n+package os\n+\n+// From DragonFly's <sys/sysctl.h>\n+const (\n+ ... | 2020-10-28T11:44:26 |
huggingface/transformers | 82a1fc7256bf27f83aec3a93543b6d156add09cf | 5e89b335ab088b52e5dbb50c5368d944e9052ed4 | Fix return_dict in encodec (#31646)
* fix: use return_dict parameter
* fix: type checks
* fix: unused imports
* update: one-line if else
* remove: recursive check | [
{
"path": "src/transformers/models/encodec/modeling_encodec.py",
"patch": "@@ -729,7 +729,7 @@ def decode(\n Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple.\n \n \"\"\"\n- return_dict = return_dict or self.config.return_dict\n+ return_dict = ... | 2024-06-28T11:18:01 |
nodejs/node | 5d35a1a3389c0e93be44a7458bc35a2a3cc9fb02 | 5bc779240ba4bd05589d5359111a231cebbccbe6 | deps: cherry-pick c690f54d95802 from V8 upstream
Original commit message:
[platform] Add TaskRunner to the platform API
With the existing platform API it is not possible to post foreground
tasks from background tasks. This is, however, required to implement
asynchronous compilation for WebAssembly. W... | [
{
"path": "deps/v8/include/v8-platform.h",
"patch": "@@ -36,6 +36,51 @@ class IdleTask {\n virtual void Run(double deadline_in_seconds) = 0;\n };\n \n+/**\n+ * A TaskRunner allows scheduling of tasks. The TaskRunner may still be used to\n+ * post tasks after the isolate gets destructed, but these tasks ma... | 2017-11-19T13:35:59 |
ggml-org/llama.cpp | f071ce67c983649e8750b03e5ea66deced34d757 | 4065c1a3a62ae0962d682e775e7f2963d8c9bf5c | ci : add action for finding duplicate issues (#20756)
* ci : add action for finding duplicates issues
* cont : gen info
* cont : formatting
* cont : fix
* cont : instructions
* cont : bump checkout action
Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com>
---------
Co-authored-by: Sigbjørn Skjæret ... | [
{
"path": ".github/workflows/ai-issues.yml",
"patch": "@@ -0,0 +1,71 @@\n+name: AI review (issues)\n+\n+on:\n+ issues:\n+ types: [opened]\n+\n+jobs:\n+ find-duplicates:\n+ if: github.event.action == 'opened'\n+ runs-on: [self-hosted, opencode]\n+\n+ permissions:\n+ contents: read\n+ ... | 2026-03-19T14:17:37 |
huggingface/transformers | 5e89b335ab088b52e5dbb50c5368d944e9052ed4 | 0142aab7f8546bc5617832740c560ab1caf2186c | Fix Gemma2 4d attention mask (#31674)
Update modeling_gemma2.py
Co-authored-by: Arthur <48595927+ArthurZucker@users.noreply.github.com> | [
{
"path": "src/transformers/models/gemma2/modeling_gemma2.py",
"patch": "@@ -629,11 +629,13 @@ def forward(\n if (\n self.config._attn_implementation != \"flash_attention_2\" and self.is_sliding and attention_mask is not None\n ): # efficient SDPA and no padding\n- at... | 2024-06-28T06:20:30 |
golang/go | d4c1ad882973e407ff85b977f4ce5b9435451190 | 421d4e72de802ed65cb38317660654771cfb13e9 | database/sql: fix tx stmt deadlock when rollback
Tx acquires tx.closemu W-lock and then acquires stmt.closemu.W-lock
to fully close the transaction and associated prepared statement.
Stmt query and execution run in reverse ways - acquires
stmt.closemu.R-lock and then acquires tx.closemu.R-lock to grab tx
connection, w... | [
{
"path": "src/database/sql/sql.go",
"patch": "@@ -2087,10 +2087,10 @@ func (tx *Tx) isDone() bool {\n // that has already been committed or rolled back.\n var ErrTxDone = errors.New(\"sql: transaction has already been committed or rolled back\")\n \n-// closeLocked returns the connection to the pool and\n+... | 2020-08-24T14:04:17 |
ollama/ollama | 7b59d1770fdcece7ee660dc0c0f1752c7b773e2f | 95ead8ffba312524c233063f89635953c44f2a73 | Fix relative path lookup | [
{
"path": "gpu/assets.go",
"patch": "@@ -40,7 +40,7 @@ func PayloadsDir() (string, error) {\n \t\t\t}\n \n \t\t\tvar paths []string\n-\t\t\tfor _, root := range []string{appExe, cwd} {\n+\t\t\tfor _, root := range []string{filepath.Dir(appExe), cwd} {\n \t\t\t\tpaths = append(paths,\n \t\t\t\t\tfilepath.Joi... | 2024-04-29T23:00:08 |
nodejs/node | c4a5de51e936940c9abce7761f2856ebf9202330 | 5d7f5c16b3ae49325b7bb0a79eaba3e945585ced | inspector: no async tracking for promises
`Promise` instances are already tracked by V8 itself.
This fixes `sequential/test-inspector-async-stack-traces-promise-then`
in debug mode (it previously crashed because our tracking
and the V8 tracking were not properly nested).
PR-URL: https://github.com/nodejs/node/pull/17... | [
{
"path": "lib/internal/inspector_async_hook.js",
"patch": "@@ -16,22 +16,33 @@ const hook = createHook({\n // in https://github.com/nodejs/node/pull/13870#discussion_r124515293,\n // this should be fine as long as we call asyncTaskCanceled() too.\n const recurring = true;\n- inspector.asyncT... | 2017-11-18T15:37:07 |
ggml-org/llama.cpp | 4065c1a3a62ae0962d682e775e7f2963d8c9bf5c | 1e645345702154ba4813d3d9bbdbd97718de82c0 | Server becomes the source of truth for sampling parameter defaults (#20558)
* webui: make server the source of truth for sampling defaults
* webui: fix Custom badge for sampling parameters
* webui: log user overrides after server sync
* chore: update webui build output
* fix: Default values for sampling settings c... | [
{
"path": "tools/server/webui/src/lib/components/app/chat/ChatSettings/ChatSettingsFields.svelte",
"patch": "@@ -5,9 +5,12 @@\n \timport Label from '$lib/components/ui/label/label.svelte';\n \timport * as Select from '$lib/components/ui/select';\n \timport { Textarea } from '$lib/components/ui/textarea';\n-... | 2026-03-19T12:20:39 |
huggingface/transformers | 1c68f2cafb4ca54562f74b66d1085b68dd6682f5 | 464aa74659b9711d2d64159ab82e1c49fa739fb7 | [HybridCache] Fix `get_seq_length` method (#31661)
* fix gemma2
* handle in generate | [
{
"path": "src/transformers/cache_utils.py",
"patch": "@@ -1083,7 +1083,7 @@ def get_max_length(self) -> Optional[int]:\n # no matter how long the sentence is\n return self.max_cache_len\n \n- def get_seq_length(self, layer_idx: Optional[int] = 0) -> int:\n+ def get_seq_length(self, la... | 2024-06-27T17:40:40 |
golang/go | c95bd2e6a99ab06efadb265bf42bbaf8d964904f | 02335cf4131f4eb1869f50f906e993676f7f414a | cmd/compile: optimize generated struct/array equality code
Use a standard "not-equal" label that we can jump to when we
detect that the arguments are not equal. This prevents the
recombination that was noticed in #39428.
Fixes #39428
Change-Id: Ib7c6b3539f4f6046043fd7148f940fcdcab70427
Reviewed-on: https://go-review... | [
{
"path": "src/cmd/compile/internal/gc/alg.go",
"patch": "@@ -529,6 +529,10 @@ func geneq(t *types.Type) *obj.LSym {\n \tfn := dclfunc(sym, tfn)\n \tnp := asNode(tfn.Type.Params().Field(0).Nname)\n \tnq := asNode(tfn.Type.Params().Field(1).Nname)\n+\tnr := asNode(tfn.Type.Results().Field(0).Nname)\n+\n+\t//... | 2020-09-16T17:13:37 |
ggml-org/llama.cpp | b486c17b3e8bee4340b4095075cefc8415d92068 | 1b9bbaa35716978dc38eb838bfe1af094ac7c1d4 | convert : support is_causal hyperparameter (#20746)
* convert : support is_causal hyperparameter
Check for the `is_causal` attribute in the Hugging Face model configuration and include it in the GGUF metadata.
* Update convert_hf_to_gguf.py
Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com>
* style: fi... | [
{
"path": "convert_hf_to_gguf.py",
"patch": "@@ -1062,6 +1062,10 @@ def set_gguf_parameters(self):\n self.gguf_writer.add_head_count_kv(n_head_kv)\n logger.info(f\"gguf: key-value head count = {n_head_kv}\")\n \n+ if self.hparams.get(\"is_causal\") is False:\n+ self... | 2026-03-19T10:41:11 |
huggingface/transformers | e44b878c0252ac1c841afcd68dd873c7fe307289 | 75a6319864225b8350c31b623ea2c73c23012a40 | Fix float out of range in owlvit and owlv2 when using FP16 or lower precision (#31657) | [
{
"path": "src/transformers/models/owlv2/modeling_owlv2.py",
"patch": "@@ -1276,7 +1276,7 @@ def forward(\n if query_mask.ndim > 1:\n query_mask = torch.unsqueeze(query_mask, dim=-2)\n \n- pred_logits = torch.where(query_mask == 0, -1e6, pred_logits)\n+ pred... | 2024-06-27T17:07:33 |
ollama/ollama | 95ead8ffba312524c233063f89635953c44f2a73 | 7aa08a77caf0121b9c9be8df1922b5fb8679d377 | Restart server on failure when running Windows app (#3985)
* app: restart server on failure
* fix linter
* address comments
* refactor log directory creation to be where logs are written
* check all log dir creation errors | [
{
"path": "app/lifecycle/server.go",
"patch": "@@ -43,37 +43,36 @@ func getCLIFullPath(command string) string {\n \treturn command\n }\n \n-func SpawnServer(ctx context.Context, command string) (chan int, error) {\n-\tdone := make(chan int)\n-\n-\tlogDir := filepath.Dir(ServerLogFile)\n-\t_, err := os.Stat(... | 2024-04-29T14:07:52 |
golang/go | 02335cf4131f4eb1869f50f906e993676f7f414a | 2414e1f17b2f82be10110513b17cdfa80c9937c6 | runtime: move TestNeedmDeadlock to crash_cgo_test.go
It requires cgo. Also, skip the test on windows and plan9.
For #42207
Change-Id: I8522773f93bc3f9826506a41a08b86a083262e31
Reviewed-on: https://go-review.googlesource.com/c/go/+/265778
Trust: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@go... | [
{
"path": "src/runtime/crash_cgo_test.go",
"patch": "@@ -600,3 +600,16 @@ func TestEINTR(t *testing.T) {\n \t\tt.Fatalf(\"want %s, got %s\\n\", want, output)\n \t}\n }\n+\n+// Issue #42207.\n+func TestNeedmDeadlock(t *testing.T) {\n+\tswitch runtime.GOOS {\n+\tcase \"plan9\", \"windows\":\n+\t\tt.Skipf(\"no... | 2020-10-28T04:05:13 |
ggml-org/llama.cpp | 1b9bbaa35716978dc38eb838bfe1af094ac7c1d4 | 07feeaa92e61d55605ca0b2be0247daeb26790b5 | common : fix gpt-oss content removal (#20745) | [
{
"path": "common/chat.cpp",
"patch": "@@ -936,7 +936,9 @@ static common_chat_params common_chat_params_init_gpt_oss(const common_chat_temp\n for (auto msg : inputs.messages) {\n if (msg.contains(\"reasoning_content\") && msg.at(\"reasoning_content\").is_string()) {\n msg[\"thinking\... | 2026-03-19T10:40:39 |
nodejs/node | 69e6c5a212622ec15b8c2cf904480b6582c6c3a5 | 2ba93f6ed7c96684ed7aeaa32db59ae8e5885f03 | http2: major update to internals
This update does several significant things:
1. It eliminates the base Nghttp2* classes and folds those
in to node::http2::Http2Session and node::http2::Http2Stream
2. It makes node::http2::Http2Stream a StreamBase instance and
sends that out to JS-land to act as the [kHandle] f... | [
{
"path": "doc/api/errors.md",
"patch": "@@ -908,6 +908,16 @@ limit.\n A message payload was specified for an HTTP response code for which a payload is\n forbidden.\n \n+<a id=\"ERR_HTTP2_PING_CANCEL\"></a>\n+### ERR_HTTP2_PING_CANCEL\n+\n+An HTTP/2 ping was cancelled.\n+\n+<a id=\"ERR_HTTP2_PING_LENGTH\"><... | 2017-11-15T18:55:31 |
huggingface/transformers | 75a6319864225b8350c31b623ea2c73c23012a40 | 727eea4ab0aa50b1b79df3de37640fd10010fd71 | Fix post gemma merge (#31660)
* nit
* toctree issue
* protect gemma2 tests as well
* sdpa supported | [
{
"path": "docs/source/en/_toctree.yml",
"patch": "@@ -382,6 +382,8 @@\n title: Fuyu\n - local: model_doc/gemma\n title: Gemma\n+ - local: model_doc/gemma2\n+ title: Gemma2\n - local: model_doc/openai-gpt\n title: GPT\n - local: model_doc/gpt_neo",
"... | 2024-06-27T15:51:42 |
ollama/ollama | 586672f4906c310d6321d42176b32bf9430fa031 | b03408de741a9e6b22025216a15657b2cb4fbfe5 | fix copying model to itself (#4019) | [
{
"path": "server/images.go",
"patch": "@@ -710,6 +710,10 @@ func CopyModel(src, dst model.Name) error {\n \t\treturn model.Unqualified(src)\n \t}\n \n+\tif src.Filepath() == dst.Filepath() {\n+\t\treturn nil\n+\t}\n+\n \tmanifests, err := GetManifestPath()\n \tif err != nil {\n \t\treturn err",
"additi... | 2024-04-29T03:47:49 |
golang/go | 2414e1f17b2f82be10110513b17cdfa80c9937c6 | 7be9158ce50e8f9f9a9a624874b8e4f3ceb07f44 | cmd/compile: make sure field offset is aligned for structure splitting
Always use the aligned form -- the old code sometimes could
at least nominally use a misaligned field in an SSA-able struct,
even if not actually.
Fixes #42181.
Change-Id: Ibdce0985f9349da70921a37423054b85ee4200d0
Reviewed-on: https://go-review.g... | [
{
"path": "src/cmd/compile/internal/gc/ssa.go",
"patch": "@@ -6978,15 +6978,10 @@ func (e *ssafn) SplitInt64(name ssa.LocalSlot) (ssa.LocalSlot, ssa.LocalSlot) {\n \n func (e *ssafn) SplitStruct(name ssa.LocalSlot, i int) ssa.LocalSlot {\n \tst := name.Type\n-\tft := st.FieldType(i)\n-\tvar offset int64\n-\... | 2020-10-26T23:06:39 |
ggml-org/llama.cpp | 3fee84e15659a2b4677200ddd1cca482839bf904 | 811397745e161563f118944ae6fab61ebcd1707e | cmake : fix build warning when kleidiai is enabled (#20457)
* cmake : fix build warning when kleidiai is enabled
* remove LLAMA_ARG_THREADS from KleidiAI backend | [
{
"path": "ggml/src/ggml-cpu/CMakeLists.txt",
"patch": "@@ -570,24 +570,34 @@ function(ggml_add_cpu_backend_variant_impl tag_name)\n set(KLEIDIAI_DOWNLOAD_URL \"https://github.com/ARM-software/kleidiai/archive/refs/tags/${KLEIDIAI_COMMIT_TAG}.tar.gz\")\n set(KLEIDIAI_ARCHIVE_MD5 \"540490375... | 2026-03-19T08:14:48 |
nodejs/node | 5df0b9e0bf16e0103cf371d34c7a2eb37bbb3739 | 9cf352553576ec3cf003a003da7e236772cd245e | src: remove unused variable
This fixes a compiler warning.
PR-URL: https://github.com/nodejs/node/pull/17150
Reviewed-By: Michael Dawson <michael_dawson@ca.ibm.com>
Reviewed-By: Anatoli Papirovski <apapirovski@mac.com>
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: Tobias Nießen <tniessen@tnie.de>
Revi... | [
{
"path": "src/node.cc",
"patch": "@@ -887,7 +887,6 @@ void SetupDomainUse(const FunctionCallbackInfo<Value>& args) {\n env->set_using_domains(true);\n \n HandleScope scope(env->isolate());\n- Local<Object> process_object = env->process_object();\n \n CHECK(args[0]->IsArray());\n env->set_domains_s... | 2017-11-20T14:29:35 |
huggingface/transformers | 0cf60f13ab1c857c17fc3fb127129048c93bf06c | 4aa17d00690b7f82c95bb2949ea57e22c35b4336 | Add gemma 2 (#31659)
* inital commit
* Add doc
* protect?
* fixup stuffs
* update tests
* fix build documentation
* mmmmmmm config attributes
* style
* nit
* uodate
* nit
* Fix docs
* protect some stuff
---------
Co-authored-by: Lysandre <lysandre@huggingface.co> | [
{
"path": "docs/source/en/index.md",
"patch": "@@ -145,6 +145,7 @@ Flax), PyTorch, and/or TensorFlow.\n | [Funnel Transformer](model_doc/funnel) | ✅ | ✅ | ❌ |\n | [Fuyu](model_doc/fuyu) ... | 2024-06-27T15:36:19 |
ggml-org/llama.cpp | 7f2cbd9a4df77a2ce10f31d69f07d4fb75eabc07 | 509a31d00f87804a466dee983a30ce29aa0441b1 | CANN: handle in-place ROPE on non-contiguous f32 tensors (#20274)
RotaryPositionEmbedding on CANN fails when src and dst share the same
non-contiguous buffer (inplace + view), because the operator overwrites
source data before it is fully read.
Add a branch that detects this case and uses contiguous temporary
buffers... | [
{
"path": "ggml/src/ggml-cann/aclnn_ops.cpp",
"patch": "@@ -2943,6 +2943,27 @@ void ggml_cann_rope(ggml_backend_cann_context & ctx, ggml_tensor * dst) {\n // Rotate full tensor (no tail), using trans tensors\n GGML_CANN_CALL_ACLNN_OP(ctx, RotaryPositionEmbedding, acl_src_trans_tensor.get(), ... | 2026-03-19T06:05:01 |
nodejs/node | 804eb3cd734176f123142f0db496279cb2f645fd | f526deb0407412e089382900e3fc4560eea3a938 | src: remove process._debugPause()
This method is undocumented and depends on a V8 API that is slated for
removal. The inspector and node-inspect don't use it and I could find
no third-party code that depends on it. Remove it.
PR-URL: https://github.com/nodejs/node/pull/17060
Reviewed-By: Anna Henningsen <anna@addal... | [
{
"path": "src/node.cc",
"patch": "@@ -67,7 +67,6 @@\n #if NODE_USE_V8_PLATFORM\n #include \"libplatform/libplatform.h\"\n #endif // NODE_USE_V8_PLATFORM\n-#include \"v8-debug.h\"\n #include \"v8-profiler.h\"\n #include \"zlib.h\"\n \n@@ -2931,7 +2930,6 @@ static void DebugPortSetter(Local<Name> property,\... | 2017-11-15T21:02:08 |
golang/go | 7be9158ce50e8f9f9a9a624874b8e4f3ceb07f44 | 368c40116434532dc0b53b72fa04788ca6742898 | cmd/link: remove all constants of elf
Use debug/elf instead.
Related:
CL 252478
CL 265317
Change-Id: If63b0458d9a6e825b40616bfb7a5a2c2e32402b4
Reviewed-on: https://go-review.googlesource.com/c/go/+/265318
Trust: Meng Zhuo <mzh@golangcn.org>
Run-TryBot: Meng Zhuo <mzh@golangcn.org>
TryBot-Result: Go Bot <gobot@golang... | [
{
"path": "src/cmd/link/internal/amd64/asm.go",
"patch": "@@ -354,7 +354,7 @@ func adddynrel(target *ld.Target, ldr *loader.Loader, syms *ld.ArchSyms, s loade\n \t\t\trela := ldr.MakeSymbolUpdater(syms.Rela)\n \t\t\trela.AddAddrPlus(target.Arch, s, int64(r.Off()))\n \t\t\tif r.Siz() == 8 {\n-\t\t\t\trela.Ad... | 2020-10-27T06:18:13 |
ollama/ollama | d6e3b64582d93dfa0c836965b70d66c8487c2f8b | 114c932a8e872846fc714353c65d041feb886027 | Fix concurrency for CPU mode
Prior refactoring passes accidentally removed the logic to bypass VRAM
checks for CPU loads. This adds that back, along with test coverage.
This also fixes loaded map access in the unit test to be behind the mutex which was
likely the cause of various flakes in the tests. | [
{
"path": "server/sched.go",
"patch": "@@ -149,6 +149,14 @@ func (s *Scheduler) processPending(ctx context.Context) {\n \t\t\t\t\t\tbreak\n \t\t\t\t\t}\n \n+\t\t\t\t\t// If we're CPU only mode, just limit by loadedMax above\n+\t\t\t\t\t// TODO handle system memory exhaustion\n+\t\t\t\t\tif (len(gpus) == 1 &... | 2024-04-28T20:40:31 |
huggingface/transformers | be50a0338b9d7b76448fcc9c5046a78118a4d968 | 3a028101e91069b51629f5e74096ae78e490022b | change anchor_image_size None for compatibility (#31640)
* change anchor_image_size None for compatibility
* make fix-copies | [
{
"path": "src/transformers/models/rt_detr/configuration_rt_detr.py",
"patch": "@@ -118,8 +118,8 @@ class RTDetrConfig(PretrainedConfig):\n Scale or magnitude of noise to be added to the bounding boxes.\n learn_initial_query (`bool`, *optional*, defaults to `False`):\n Indica... | 2024-06-27T11:36:55 |
ggml-org/llama.cpp | 509a31d00f87804a466dee983a30ce29aa0441b1 | ea01d196d7d4fcf1aa67496d493bcad9a9087360 | ggml-webgpu: Update the `RMS_NORM` preprocessor and add `L2_NORM` (#20665)
* Update the preprocessor of RMS_NORM and add L2_NORM.
* Fix the name of rms_norm to row_norm. | [
{
"path": "docs/ops.md",
"patch": "@@ -62,7 +62,7 @@ Legend:\n | HARDSWISH | ❌ | ✅ | ✅ | 🟡 | 🟡 | ❌ | ✅ | 🟡 | ✅ | ❌ | ❌ |\n | IM2COL | ❌ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | ❌ | ❌ | ❌ |\n | IM2COL_3D | ❌ | ❌ | ✅ | ✅ | ❌ | ❌ | ❌ | ✅ | ❌ | ❌ |... | 2026-03-19T04:08:59 |
nodejs/node | f526deb0407412e089382900e3fc4560eea3a938 | 7dc35e937d6b3cf57e6184dfc3d54633e5b69082 | src: inspector context name = program title + pid
Report (for example) "node[1337]" as the human-readable name rather
than the more generic and less helpful "Node.js Main Context."
While not perfect yet, it should be an improvement to people that
debug multiple processes from DevTools, VS Code, etc.
PR-URL: https://... | [
{
"path": "src/inspector_agent.cc",
"patch": "@@ -302,7 +302,8 @@ class NodeInspectorClient : public V8InspectorClient {\n : env_(env), platform_(platform), terminated_(false),\n running_nested_loop_(false) {\n client_ = V8Inspector::create(env->isolate(), this);\n- contextCreated(env->... | 2017-11-20T22:37:50 |
golang/go | 368c40116434532dc0b53b72fa04788ca6742898 | 94f3762462a999bfc5491c8d1b892110651e23d6 | runtime: block signals in needm before allocating M
Otherwise, if a signal occurs just after we allocated the M,
we can deadlock if the signal handler needs to allocate an M
itself.
Fixes #42207
Change-Id: I76f44547f419e8b1c14cbf49bf602c6e645d8c14
Reviewed-on: https://go-review.googlesource.com/c/go/+/265759
Trust: ... | [
{
"path": "src/runtime/crash_unix_test.go",
"patch": "@@ -358,3 +358,12 @@ func TestSignalM(t *testing.T) {\n \t\tt.Fatalf(\"signal sent to M %d, but received on M %d\", want, got)\n \t}\n }\n+\n+// Issue #42207.\n+func TestNeedmDeadlock(t *testing.T) {\n+\toutput := runTestProg(t, \"testprogcgo\", \"NeedmD... | 2020-10-27T23:09:40 |
huggingface/transformers | 11138ca013add0041753aab276e83261c38b08ac | c9f191a0b7f391594901a960671f2a199122ef48 | [`Llama`] Conversion: fix and simplify the script! (#31591)
* fix and simplify the script!
* add co-author
---------
Co-authored-by: crackalamoo <crackalamoo@users.noreply.github.com> | [
{
"path": "src/transformers/models/llama/convert_llama_weights_to_hf.py",
"patch": "@@ -105,21 +105,18 @@ def write_json(text, path):\n def write_model(\n model_path,\n input_base_path,\n- model_size,\n+ model_size=None,\n safe_serialization=True,\n llama_version=1,\n vocab_size=No... | 2024-06-27T10:35:19 |
ggml-org/llama.cpp | 07ba6d275b0f5c138c72f75d7f3df2661f17c27a | 6729d4920c7509f0d110f114a9652793b5fe668a | CANN: support flash attention for head dim not multiple of 16, fix ALiBi slope offset (#20031)
- Allow FLASH_ATTN_EXT when head dimension D is not a multiple of 16 by
padding Q/K/V to D_padded = GGML_PAD(D, 16), running FusedInferAttentionScoreV2,
then slicing the output back to D (ggml-cann.cpp + aclnn_ops.cpp).
... | [
{
"path": "ggml/src/ggml-cann/aclnn_ops.cpp",
"patch": "@@ -1544,8 +1544,8 @@ static void aclnn_get_slope(ggml_backend_cann_context & ctx,\n end = 2 * ((n_head - 1) - n_head_log2) + 1;\n step = 2;\n count = n_head - n_head_log2;\n- aclnn_get_slope_inner(ctx, (char *) slope_... | 2026-03-19T03:02:42 |
nodejs/node | 5c7b1ecbceeb32d947bea31c1256b1a9a42b8cba | 52010553d3c37b9d0b75d4c3ca9e63956a59a047 | test: make debugging of inspector-port-zero easier
If the process was killed, then the exit code will be null, in which
case knowing the signal is really helpful.
PR-URL: https://github.com/nodejs/node/pull/16685
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By:... | [
{
"path": "test/sequential/test-inspector-port-zero.js",
"patch": "@@ -34,7 +34,10 @@ function test(arg, port = '') {\n };\n proc.stdout.on('close', mustCall(() => onclose()));\n proc.stderr.on('close', mustCall(() => onclose()));\n- proc.on('exit', mustCall((exitCode) => assert.strictEqual(e... | 2017-11-02T10:56:27 |
huggingface/transformers | c9f191a0b7f391594901a960671f2a199122ef48 | dc76e9fa7f0d19ff7cfc33bd3a22acd7df167fce | Fix ONNX exports for Optimum compatible models (#31311)
* fixed models
* format with bumped ruff version on my local
* fix copies
* add tracing checks
* format
* Update src/transformers/utils/generic.py
Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com>
* format
* style... | [
{
"path": "src/transformers/models/clap/modeling_clap.py",
"patch": "@@ -37,6 +37,7 @@\n add_start_docstrings_to_model_forward,\n logging,\n replace_return_docstrings,\n+ torch_int,\n )\n from .configuration_clap import ClapAudioConfig, ClapConfig, ClapTextConfig\n \n@@ -590,8 +591,10 @@ def ... | 2024-06-27T09:46:36 |
golang/go | 94f3762462a999bfc5491c8d1b892110651e23d6 | 49b017fe59bf628795f2c4fdbcb5db942e865fa9 | cmd/go: add -include to cgo whitelist
Fixes #39988.
Change-Id: Ia6f5b73e6508f27e3badbcbd29dbeadffd55a932
Reviewed-on: https://go-review.googlesource.com/c/go/+/240739
Trust: Tobias Klauser <tobias.klauser@gmail.com>
Run-TryBot: Tobias Klauser <tobias.klauser@gmail.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewe... | [
{
"path": "src/cmd/go/internal/work/security.go",
"patch": "@@ -132,6 +132,7 @@ var validCompilerFlagsWithNextArg = []string{\n \t\"-U\",\n \t\"-I\",\n \t\"-framework\",\n+\t\"-include\",\n \t\"-isysroot\",\n \t\"-isystem\",\n \t\"--sysroot\",",
"additions": 1,
"deletions": 0,
"language": "Go"
... | 2020-07-02T00:48:37 |
ggml-org/llama.cpp | 78d550b541a12eb473f1bca4cf9ac6920ebdb42b | 4efd326e7107f3c6a84d23455ac620e7220079ec | ggml-cpu/x86: fix unused changemask warning in repack (#20692) | [
{
"path": "ggml/src/ggml-cpu/arch/x86/repack.cpp",
"patch": "@@ -531,7 +531,6 @@ static void gemv_q4_b32_8x8_q8_0_lut_avx(int n, float * GGML_RESTRICT s, size_t\n \n UNUSED(bs);\n \n- __m128i changemask = _mm_set_epi8(15, 14, 7, 6, 13, 12, 5, 4, 11, 10, 3, 2, 9, 8, 1, 0);\n __m256i finalpermutema... | 2026-03-18T16:45:06 |
huggingface/transformers | dc76e9fa7f0d19ff7cfc33bd3a22acd7df167fce | 1de7dc7403b3b89ec421d43a8c9ee245211a61f6 | Generation: past kv can be None (#31051)
* fix
* better | [
{
"path": "src/transformers/generation/utils.py",
"patch": "@@ -1394,7 +1394,7 @@ def _get_initial_cache_position(self, input_ids, model_kwargs):\n return model_kwargs\n \n past_length = 0\n- if \"past_key_values\" in model_kwargs:\n+ if model_kwargs.get(\"past_key_values\"... | 2024-06-27T04:55:33 |
nodejs/node | 52010553d3c37b9d0b75d4c3ca9e63956a59a047 | e51216d722ca43b2db7f79c0c23e917f5b85bd3d | test: replace assert.throws w/ common.expectsError
PR-URL: https://github.com/nodejs/node/pull/17091
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Michaël Zasso <targos@protonmail.com>
Reviewed-By: Luigi Pinca <luigipinca@gmail.com>
Reviewed-By: Gireesh Punathil <gpunathi@in.ibm.com> | [
{
"path": "test/parallel/test-next-tick-errors.js",
"patch": "@@ -20,7 +20,7 @@\n // USE OR OTHER DEALINGS IN THE SOFTWARE.\n \n 'use strict';\n-require('../common');\n+const common = require('../common');\n const assert = require('assert');\n \n const order = [];\n@@ -42,11 +42,15 @@ process.nextTick(funct... | 2017-11-17T06:46:14 |
golang/go | 49b017fe59bf628795f2c4fdbcb5db942e865fa9 | 40d1ec588f492f63637b59e933d6c682d66659a4 | runtime: handle signal 34 for musl setgid
It has been observed that setgid hangs when using cgo with musl.
This fix ensures that signal 34 gets handled in an appropriate way,
like signal 33 when using glibc.
Fixes #39343
Change-Id: I89565663e2c361f62cbccfe80aaedf290bd58d57
Reviewed-on: https://go-review.googlesource... | [
{
"path": "src/runtime/sigtab_linux_generic.go",
"patch": "@@ -45,7 +45,7 @@ var sigtable = [...]sigTabT{\n \t/* 31 */ {_SigThrow, \"SIGSYS: bad system call\"},\n \t/* 32 */ {_SigSetStack + _SigUnblock, \"signal 32\"}, /* SIGCANCEL; see issue 6997 */\n \t/* 33 */ {_SigSetStack + _SigUnblock, \"signal 33\"},... | 2020-06-04T20:11:56 |
ollama/ollama | ed5fb088c4d7f57ea3ccf629e36ecb2e857e22eb | 11d83386a5761b31f78e060054483c6f0c70dc7e | Fix target in gen_windows.ps1 | [
{
"path": "llm/generate/gen_windows.ps1",
"patch": "@@ -26,7 +26,6 @@ function amdGPUs {\n $GPU_LIST -join ';'\n }\n \n-$script:cmakeTargets = @(\"ollama_llama_server\")\n \n function init_vars {\n if (!$script:SRC_DIR) {\n@@ -35,6 +34,9 @@ function init_vars {\n if (!$script:llamacppDir) {\n ... | 2024-04-26T22:10:42 |
ggml-org/llama.cpp | f4049ad735a8e7b3b14762a8e3bb79784b67da44 | 5e8910a0db1a577e102f63313eb23f285a98d205 | tests : fix test-jinja-py Windows failures by bypassing command-line args [no ci] (#20483)
* Fix errors occurring on Windows
* Reverted fix
#20365 will take care of CRLF isue
* Changed to write to directly to stdin
* Prevent fclose to happen twice | [
{
"path": "tests/test-jinja.cpp",
"patch": "@@ -1897,8 +1897,9 @@ import sys\n from datetime import datetime\n from jinja2.sandbox import SandboxedEnvironment\n \n-tmpl = json.loads(sys.argv[1])\n-vars_json = json.loads(sys.argv[2])\n+merged_input = json.loads(sys.stdin.buffer.read().decode(\"utf-8\"))\n+tm... | 2026-03-18T09:43:31 |
golang/go | 70e022e4a83dc996ac4f108e811fbc399ad5565b | 5d3666e1a48d0976718c75dddc2ef0232be835d8 | time: fix LoadLocationFromTZData with slim tzdata
The extend information of a time zone file with last transition < now
could result in a wrong cached zone because it used the zone of the
last transition.
This could lead to wrong zones in systems with slim zoneinfo.
Fixes #42216
Change-Id: I7c57c35b5cfa58482ac7925b... | [
{
"path": "src/time/zoneinfo_read.go",
"patch": "@@ -323,17 +323,26 @@ func LoadLocationFromTZData(name string, data []byte) (*Location, error) {\n \t\tif tx[i].when <= sec && (i+1 == len(tx) || sec < tx[i+1].when) {\n \t\t\tl.cacheStart = tx[i].when\n \t\t\tl.cacheEnd = omega\n+\t\t\tzoneIdx := tx[i].index... | 2020-10-26T12:44:44 |
nodejs/node | 289fcb05be0efe5f57c742af7113547a4af94c69 | 07a4fa3dbefdc7f740a5a281f713998e71aab8f3 | build: do not build doc in source tarball
PR-URL: https://github.com/nodejs/node/pull/17100
Fixes: https://github.com/nodejs/node/issues/16650
Reviewed-By: Refael Ackermann <refack@gmail.com> | [
{
"path": "Makefile",
"patch": "@@ -531,49 +531,68 @@ endif\n # generated .html files\n DOCS_ANALYTICS ?=\n \n+apidoc_dirs = out/doc out/doc/api out/doc/api/assets\n apidoc_sources = $(wildcard doc/api/*.md)\n-apidocs_html = $(apidoc_dirs) $(apiassets) $(addprefix out/,$(apidoc_sources:.md=.html))\n-apidocs... | 2017-11-17T17:19:22 |
ollama/ollama | f81f30811878ee3b59deaa2319eec489fafb39ef | 7fea1ecdf68103cdc329b7f8c65904cd54c3e56d | fix gemma, command-r layer weights | [
{
"path": "llm/memory.go",
"patch": "@@ -102,10 +102,14 @@ func EstimateGPULayers(gpus []gpu.GpuInfo, ggml *GGML, projectors []string, opts\n \tlayers := ggml.Tensors().Layers()\n \n \tvar memoryLayerOutput uint64\n-\tfor k, v := range layers {\n-\t\tif k == \"output\" || k == \"output_norm\" {\n-\t\t\tmemo... | 2024-04-26T22:00:54 |
ggml-org/llama.cpp | 5e8910a0db1a577e102f63313eb23f285a98d205 | fe00a84b4b972b3172b2c6b880954d81da532ca4 | common : rework gpt-oss parser (#20393)
* common : rework gpt-oss parser
* cont : fix gpt-oss tests
* cont : add structured output test
* cont : rename final to final_msg | [
{
"path": "common/chat.cpp",
"patch": "@@ -933,17 +933,12 @@ static common_chat_params common_chat_params_init_gpt_oss(const common_chat_temp\n \n // Copy reasoning to the \"thinking\" field as expected by the gpt-oss template\n auto adjusted_messages = json::array();\n- for (const auto & msg : i... | 2026-03-18T09:41:25 |
rust-lang/rust | f0dec714f3b6b6feaa5a0f40685e1e520636dff5 | 03eb45452305f2d52348279d0caa5fc1f12c438d | Make some invalid codegen attr errors structured/translatable | [
{
"path": "compiler/rustc_codegen_ssa/messages.ftl",
"patch": "@@ -18,6 +18,8 @@ codegen_ssa_atomic_compare_exchange = Atomic compare-exchange intrinsic missing\n \n codegen_ssa_autodiff_without_lto = using the autodiff feature requires using fat-lto\n \n+codegen_ssa_bare_instruction_set = `#[instruction_se... | 2025-02-19T16:47:28 |
huggingface/transformers | 1de7dc7403b3b89ec421d43a8c9ee245211a61f6 | 1f9f57ab4c8c30964360a2ba697c339f6d31f03f | Skip tests properly (#31308)
* Skip tests properly
* [test_all]
* Add 'reason' as kwarg for skipTest
* [test_all] Fix up
* [test_all] | [
{
"path": "docs/source/de/testing.md",
"patch": "@@ -185,16 +185,16 @@ pytest -k \"test and ada\" tests/test_optimization.py\n \n Manchmal müssen Sie `accelerate` Tests für Ihre Modelle ausführen. Dazu fügen Sie einfach `-m accelerate_tests` zu Ihrem Befehl hinzu, wenn Sie diese Tests bei einem `OPT`-Lauf a... | 2024-06-26T20:59:08 |
ggml-org/llama.cpp | fe00a84b4b972b3172b2c6b880954d81da532ca4 | 7ab321d40db7f26f910bff51eed92fb614994229 | tests: enable kv_unified to prevent cuda oom error on rtx 2060 (#20645)
Signed-off-by: Aaron Teo <aaron.teo1@ibm.com> | [
{
"path": "tests/test-backend-sampler.cpp",
"patch": "@@ -89,6 +89,7 @@ struct test_context {\n cparams.n_batch = 512;\n cparams.samplers = configs.data();\n cparams.n_samplers = configs.size();\n+ cparams.kv_unified = true;\n \n // If n_seq_max is not specified, calcu... | 2026-03-18T09:40:22 |
nodejs/node | f841fc63ddc4808e301e30b60f666190c025f9a2 | ce6c7f5b01a9faf5b4c6561944e89ba8d42b68de | src: use std::unique_ptr in base-object-inl.h
Drive-by fix: delete superflouos nullptr in ptr.reset().
PR-URL: https://github.com/nodejs/node/pull/17079
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Timothy Gu <timothygu99@gmail.com> | [
{
"path": "src/base_object-inl.h",
"patch": "@@ -65,9 +65,8 @@ inline Environment* BaseObject::env() const {\n template <typename Type>\n inline void BaseObject::WeakCallback(\n const v8::WeakCallbackInfo<Type>& data) {\n- Type* self = data.GetParameter();\n+ std::unique_ptr<Type> self(data.GetParamet... | 2017-11-16T14:50:39 |
golang/go | 6e98406ac3e654f4df15f662f51eda46434af332 | 38af45b4cb4ce42b338ba9960419684b2c2c5e72 | [dev.typeparams] cmd/compile/internal/types2: review of initorder.go
Difference: errorf now accepts any value that implements
the poser interface in place of a position argument. All
types2 Objects implement poser.
type poser interface {
Pos() syntax.Pos
}
f=initorder.go; diff $f ../../../../go/types/$f
5c5
< packa... | [
{
"path": "src/cmd/compile/internal/types2/initorder.go",
"patch": "@@ -1,4 +1,3 @@\n-// UNREVIEWED\n // Copyright 2014 The Go Authors. All rights reserved.\n // Use of this source code is governed by a BSD-style\n // license that can be found in the LICENSE file.",
"additions": 0,
"deletions": 1,
... | 2020-10-27T18:26:54 |
rust-lang/rust | 9212e31c92aba29fac1c1e05a258abe36dd0afee | beba32cebb4fb2ef8f02c4fc898a9d31f1b03c61 | Add `unreachable_pub` to the default lints for `compiler/` crates.
And fix the new errors in the handful of crates that didn't have a
`#![warn(unreachable_pub)]`. | [
{
"path": "Cargo.toml",
"patch": "@@ -74,6 +74,7 @@ exclude = [\n # FIXME(edition_2024): Change this to `-Wrust_2024_idioms` when all of the\n # individual lints are satisfied.\n keyword_idents_2024 = \"warn\"\n+unreachable_pub = \"warn\"\n unsafe_op_in_unsafe_fn = \"warn\"\n \n [profile.release.package.rus... | 2025-03-06T03:25:44 |
huggingface/transformers | 1f9f57ab4c8c30964360a2ba697c339f6d31f03f | a3fb96a42a9ee473de61ac01a860251123042943 | Fix dtype casting in swinv2 and swinv2sr to allow non-FP32 inference (#31589)
* Fix dtype casting in modeling_swin2sr to allow non-FP32 inference
* Fix formattting
* Fix for swinv2 too
* Update src/transformers/models/swin2sr/modeling_swin2sr.py
Co-authored-by: amyeroberts <22614925+amyeroberts@users.norep... | [
{
"path": "src/transformers/models/swin2sr/modeling_swin2sr.py",
"patch": "@@ -301,6 +301,8 @@ def __init__(self, config, dim, num_heads, window_size, pretrained_window_size=[\n relative_coords_table = (\n torch.sign(relative_coords_table) * torch.log2(torch.abs(relative_coords_table) + ... | 2024-06-26T17:46:48 |
ggml-org/llama.cpp | 7ab321d40db7f26f910bff51eed92fb614994229 | 7533a7d5096894df1d9dd842856cd4b7e16ef4ba | webui: Fix duplicated messages on q param (#20715)
* fix: Remove duplicate message sending on `?q` param
* chore: update webui build output | [
{
"path": "tools/server/webui/src/routes/+page.svelte",
"patch": "@@ -57,7 +57,6 @@\n \t\t// Handle ?q= parameter - create new conversation and send message\n \t\tif (qParam !== null) {\n \t\t\tawait conversationsStore.createConversation();\n-\t\t\tawait chatStore.sendMessage(qParam);\n \t\t\tclearUrlParams... | 2026-03-18T09:32:43 |
golang/go | 5d3666e1a48d0976718c75dddc2ef0232be835d8 | b4b014465216790e01aa66f9120d03230e4aff46 | crypto/tls: document the ClientAuthType consts
Fixes #34023
Change-Id: Ib7552a8873a79a91e8d971f906c6d7283da7a80c
Reviewed-on: https://go-review.googlesource.com/c/go/+/264027
Trust: Roland Shoemaker <roland@golang.org>
Reviewed-by: Katie Hockman <katie@golang.org> | [
{
"path": "src/crypto/tls/common.go",
"patch": "@@ -294,10 +294,26 @@ func (cs *ConnectionState) ExportKeyingMaterial(label string, context []byte, le\n type ClientAuthType int\n \n const (\n+\t// NoClientCert indicates that no client certificate should be requested\n+\t// during the handshake, and if any c... | 2020-10-21T17:59:22 |
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.