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 |
|---|---|---|---|---|---|
ggml-org/llama.cpp | 70d860824a72c3daaea678b74d5051fdd46d9c38 | 080b161995218bb40bfc03a6446c9e2b4c7e81e0 | convert : fix Gemma3N, GraniteMoe and Ernie4.5Moe (#19084)
* fix Gemma3N and Ernie4.5Moe
* fix GraniteMoe | [
{
"path": "convert_hf_to_gguf.py",
"patch": "@@ -3799,7 +3799,7 @@ def modify_tensors(self, data_torch: Tensor, name: str, bid: int | None) -> Iter\n merged_name = f\"model.layers.{bid}.mlp.experts.{w_name}.weight\"\n yield from super().modify_tensors(data_torch, merg... | 2026-01-25T12:05:05 |
golang/go | 3aa09489ab3aa13a3ac78b1ff012b148ffffe367 | 4e1d812afc2ebe767face21f34e47de57f3f32a6 | cmd/go: add a '-e' flag to 'mod tidy' and 'mod vendor'
This flag, like the -e flag to 'go list', instructs the command to
make a best effort to continue in spite of errors for specific packages.
Fixes #26603
Change-Id: I5ee2f50c71870ae8ef3f9b3e5b045474adcca525
Reviewed-on: https://go-review.googlesource.com/c/go/+/2... | [
{
"path": "doc/go1.16.html",
"patch": "@@ -66,6 +66,13 @@ <h4 id=\"modules\">Modules</h4>\n TODO: write and link to tutorial or blog post\n </p>\n \n+<p><!-- golang.org/issue/26603 -->\n+ The <code>go</code> <code>mod</code> <code>vendor</code>\n+ and <code>go</code> <code>mod</code> <code>tidy</code> s... | 2020-09-18T14:23:58 |
huggingface/transformers | df475bf8e62e843bfd4f604b81696b2d950ec990 | e076953079469983b8430fdb032f8a61dc2ae5f4 | Trainer - add cache clearing and the option for batched eval metrics computation (#28769)
* Added cache clearing for GPU efficiency.
* Added cache clearing for GPU efficiency.
* Added batch_eval_metrics capability
* Ran make fixup
* Fixed bug
* Fixed whitespace issue
* Fixed outdated condition
* U... | [
{
"path": "src/transformers/trainer.py",
"patch": "@@ -327,7 +327,10 @@ class Trainer:\n inner layers, dropout probabilities etc).\n compute_metrics (`Callable[[EvalPrediction], Dict]`, *optional*):\n The function that will be used to compute metrics at evaluation. Must take ... | 2024-05-06T12:23:40 |
ollama/ollama | 39be7fdb98cdcdfb1da0753410bccfc30cfa9c0d | c2e3b891760686d163b701be564dcbcab87a8785 | fix rhel cuda install (#1321)
Co-authored-by: Cloud User <azureuser@testgpu2.hqzwom21okjenksna4y3c4ymjd.phxx.internal.cloudapp.net> | [
{
"path": "scripts/install.sh",
"patch": "@@ -217,7 +217,7 @@ fi\n \n if ! check_gpu nvidia-smi || [ -z \"$(nvidia-smi | grep -o \"CUDA Version: [0-9]*\\.[0-9]*\")\" ]; then\n case $OS_NAME in\n- centos|rhel) install_cuda_driver_yum 'rhel' $OS_VERSION ;;\n+\tcentos|rhel) install_cuda_driver_yum '... | 2023-11-29T19:55:15 |
ggml-org/llama.cpp | 080b161995218bb40bfc03a6446c9e2b4c7e81e0 | 1243f93a2de868e16a9e52af55b7ab930110c04e | completion : fix prompt cache for recurrent models (#19045) | [
{
"path": "src/llama-context.cpp",
"patch": "@@ -2559,6 +2559,7 @@ size_t llama_context::state_write_data(llama_io_write_i & io) {\n }\n }\n \n+ // [TAG_CONTEXT_STATE_LOGITS]\n // write logits\n {\n LLAMA_LOG_DEBUG(\"%s: - writing logits\\n\", __func__);",
"additions": 1,
... | 2026-01-25T07:12:50 |
nodejs/node | ba4a0a6f5f9f0172e182da4e9000fec8956d0689 | e79a61cf80335498c3150dafd9036de8ec6b9914 | doc: fix http2 example with rstWithCancel
Replace the non-existent method rstStreamWithCancel with rstWithCancel
PR-URL: https://github.com/nodejs/node/pull/16365
Reviewed-By: Joyee Cheung <joyeec9h3@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Anatoli Papirovski <apapirovski@mac.com>
Review... | [
{
"path": "doc/api/http2.md",
"patch": "@@ -817,7 +817,7 @@ const client = http2.connect('http://example.org:8000');\n const req = client.request({ ':path': '/' });\n \n // Cancel the stream if there's no activity after 5 seconds\n-req.setTimeout(5000, () => req.rstStreamWithCancel());\n+req.setTimeout(5000... | 2017-10-21T18:44:45 |
golang/go | fd75989f46c80d2446dd9dcefaffbebdb7f7ea87 | d42b32e321fa5c5d2c93b2ad22d48e804c9f45d2 | cmd/go/internal/modget: consolidate Load entrypoints
This change replaces ImportPaths, ImportPathsQuiet, LoadALL, and
LoadVendor with a single LoadPackages function, with a LoadOpts struct
that more clearly documents the variations in behavior.
It also eliminates the cmd/go/internal/load.ImportPaths function,
which w... | [
{
"path": "src/cmd/go/internal/list/list.go",
"patch": "@@ -325,7 +325,7 @@ var (\n var nl = []byte{'\\n'}\n \n func runList(ctx context.Context, cmd *base.Command, args []string) {\n-\tmodload.LoadTests = *listTest\n+\tload.ModResolveTests = *listTest\n \twork.BuildInit()\n \tout := newTrackingWriter(os.St... | 2020-09-17T20:24:29 |
huggingface/transformers | aa64f086a2a229315e5a52050600a64872833580 | 9c772ac88883dd35166b58cc8cf10cffa3ca7844 | Fix llava next tie_word_embeddings config (#30640)
* fix llava next embedding
* add docstring
* Update src/transformers/models/llava_next/configuration_llava_next.py
Co-authored-by: NielsRogge <48327001+NielsRogge@users.noreply.github.com>
---------
Co-authored-by: NielsRogge <48327001+NielsRogge@users... | [
{
"path": "src/transformers/models/llava_next/configuration_llava_next.py",
"patch": "@@ -55,6 +55,8 @@ class LlavaNextConfig(PretrainedConfig):\n image_grid_pinpoints (`List`, *optional*, defaults to `[[336, 672], [672, 336], [672, 672], [1008, 336], [336, 1008]]`):\n A list of possible... | 2024-05-06T12:01:26 |
ollama/ollama | c2e3b891760686d163b701be564dcbcab87a8785 | cde31cb220f0d3b4b6eba75faf854a542a332384 | fix: disable ':' in tag names (#1280)
Co-authored-by: rootedbox | [
{
"path": "server/routes.go",
"patch": "@@ -416,6 +416,11 @@ func CreateModelHandler(c *gin.Context) {\n \t\treturn\n \t}\n \n+\tif strings.Count(req.Name, \":\") > 1 {\n+\t\tc.AbortWithStatusJSON(http.StatusBadRequest, gin.H{\"error\": \"':' (colon) is not allowed in tag names\"})\n+\t\treturn\n+\t}\n+\n \... | 2023-11-29T18:33:45 |
ggml-org/llama.cpp | 24bc23830313ce13080ae393309cef080f2141c2 | 16639ba2178fc1ccbae912d678d8b515dc4dc322 | llama: fix integer type consistency in split helpers (#18894)
* llama: fix integer type consistency in split helpers
* llama: apply minor style fixes
* llama: remove trailing whitespace | [
{
"path": "include/llama.h",
"patch": "@@ -1476,12 +1476,12 @@ extern \"C\" {\n /// @details Build a split GGUF final path for this chunk.\n /// llama_split_path(split_path, sizeof(split_path), \"/models/ggml-model-q4_0\", 2, 4) => split_path = \"/models/ggml-model-q4_0-00002-of-00004.gguf\... | 2026-01-25T07:10:52 |
nodejs/node | e79a61cf80335498c3150dafd9036de8ec6b9914 | f1d6b04ac95abe9ddbb6817b215bcdc6bef33ada | buffer: buffer.transcode to use internal/errors
`buffer.transcode` is still using raw TypeError. This change is to
convert it to use internal/errors.
Ref: https://github.com/nodejs/node/issues/11273
PR-URL: https://github.com/nodejs/node/pull/16352
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Luigi Pin... | [
{
"path": "lib/buffer.js",
"patch": "@@ -1562,7 +1562,8 @@ if (process.binding('config').hasIntl) {\n // Buffer instance.\n transcode = function transcode(source, fromEncoding, toEncoding) {\n if (!isUint8Array(source))\n- throw new TypeError('\"source\" argument must be a Buffer or Uint8Array'... | 2017-10-21T03:13:04 |
huggingface/transformers | 9c772ac88883dd35166b58cc8cf10cffa3ca7844 | a45c5148993a67c9fb5ac157754e68dfeba6f076 | Quantization / HQQ: Fix HQQ tests on our runner (#30668)
Update test_hqq.py | [
{
"path": "tests/quantization/hqq/test_hqq.py",
"patch": "@@ -35,7 +35,7 @@\n \n \n class HQQLLMRunner:\n- def __init__(self, model_id, quant_config, compute_dtype, device, cache_dir):\n+ def __init__(self, model_id, quant_config, compute_dtype, device, cache_dir=None):\n self.model = AutoMode... | 2024-05-06T09:33:52 |
golang/go | d42b32e321fa5c5d2c93b2ad22d48e804c9f45d2 | 53c9b9588a3a811bdf8d7ac2ff371bc2f95ed261 | runtime: add sched.lock assertions
Functions that require holding sched.lock now have an assertion.
A few places with missing locks have been fixed in this CL:
Additionally, locking is added around the call to procresize in
schedinit. This doesn't technically need a lock since the program is
still starting (thus no ... | [
{
"path": "src/runtime/proc.go",
"patch": "@@ -591,6 +591,7 @@ func schedinit() {\n \tparsedebugvars()\n \tgcinit()\n \n+\tlock(&sched.lock)\n \tsched.lastpoll = uint64(nanotime())\n \tprocs := ncpu\n \tif n, ok := atoi32(gogetenv(\"GOMAXPROCS\")); ok && n > 0 {\n@@ -599,6 +600,7 @@ func schedinit() {\n \ti... | 2020-08-21T15:51:25 |
ggml-org/llama.cpp | 9981c3013018b1e955f2e44dbe514032496d3f6e | e9fd8dcab45d6cd147874e32565923bdfd0efbdb | convert : fix conversion for inheriting models that were bypassing modify_tensors (#19064)
* Add undo_permute = False where needed
* Replace super().modify_tensors with ModelBase
* Add one more ModelBase.modify_tensors
* Update convert_hf_to_gguf.py
Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com>
*... | [
{
"path": "convert_hf_to_gguf.py",
"patch": "@@ -2736,7 +2736,7 @@ def modify_tensors(self, data_torch: Tensor, name: str, bid: int | None) -> Iter\n \n data_torch = torch.stack(datas, dim=0)\n merged_name = f\"model.layers.{bid}.mlp.experts.{w_name}.weight\"\n- ... | 2026-01-25T01:36:47 |
huggingface/transformers | a45c5148993a67c9fb5ac157754e68dfeba6f076 | 09edd77f64481cc9ce03067f4cfe74cef9d5d776 | Hotfix-change-ci (#30669)
* dmmy change
* fiux
* revert change | [
{
"path": ".circleci/create_circleci_config.py",
"patch": "@@ -152,7 +152,7 @@ def to_dict(self):\n tests = \" \".join(expanded_tests)\n \n # Each executor to run ~10 tests\n- n_executors = max(len(tests) // 10, 1)\n+ n_executors = max(len(expanded_tests) // 10,... | 2024-05-06T09:26:04 |
nodejs/node | f1d6b04ac95abe9ddbb6817b215bcdc6bef33ada | 556ebab30e171c2941deb529a42ad47301972af5 | src: use new V8 API in vm module
Remove the CopyProperties() hack in the vm module, i.e., Contextify.
Use different V8 API methods, that allow interception of
DefineProperty() and do not flatten accessor descriptors to
property descriptors.
Move many known issues to test cases. Factor out the last test in
test-vm-con... | [
{
"path": "src/node_contextify.cc",
"patch": "@@ -38,6 +38,7 @@ using v8::Function;\n using v8::FunctionCallbackInfo;\n using v8::FunctionTemplate;\n using v8::HandleScope;\n+using v8::IndexedPropertyHandlerConfiguration;\n using v8::Integer;\n using v8::Just;\n using v8::Local;\n@@ -58,13 +59,19 @@ using v... | 2017-10-17T15:27:27 |
golang/go | 23573d0ea225d4b93ccd2b946b1de121c3a6cee5 | 8860251db862a091ec70e3571f0885d232187119 | cmd/compile: clearer error when non-bool used as "||" and "&&" operand
Fixes #41500
Change-Id: I658d8921b7769b6e4288ca781cbdca5ff14a84ee
Reviewed-on: https://go-review.googlesource.com/c/go/+/255899
Trust: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: G... | [
{
"path": "src/cmd/compile/internal/gc/typecheck.go",
"patch": "@@ -630,6 +630,22 @@ func typecheck1(n *Node, top int) (res *Node) {\n \t\t\tbreak\n \t\t}\n \n+\t\t// For \"x == x && len(s)\", it's better to report that \"len(s)\" (type int)\n+\t\t// can't be used with \"&&\" than to report that \"x == x\" ... | 2020-09-21T05:00:24 |
ggml-org/llama.cpp | bb02f74c612064947e51d23269a1cf810b67c9a7 | 8f91ca54ec0b22f3ff3a495f32be8e8300638cdf | chat: fix language input for translategemma (#19052)
* chat: fix language input for translategemma
* Update common/chat.cpp
Co-authored-by: Aldehir Rojas <hello@alde.dev>
---------
Co-authored-by: Aldehir Rojas <hello@alde.dev> | [
{
"path": "common/chat.cpp",
"patch": "@@ -2659,6 +2659,10 @@ static common_chat_params common_chat_params_init_translate_gemma(const common_c\n templates_params inputs_new = inputs;\n json & messages = inputs_new.messages;\n \n+ // default to chat_template_kwargs, or en-GB if not specified\n+ ... | 2026-01-24T16:58:45 |
ollama/ollama | 424d53ac70d1960770f3bb3a7aa44afea9f2dbeb | e1a69d44c9c9adae94a54e123a7118915d6bbf4e | progress: fix bar rate | [
{
"path": "format/bytes.go",
"patch": "@@ -37,6 +37,8 @@ func HumanBytes(b int64) string {\n \tswitch {\n \tcase value >= 100:\n \t\treturn fmt.Sprintf(\"%d %s\", int(value), unit)\n+\tcase value >= 10:\n+\t\treturn fmt.Sprintf(\"%d %s\", int(value), unit)\n \tcase value != math.Trunc(value):\n \t\treturn f... | 2023-11-19T00:23:03 |
huggingface/transformers | 307f632bb2fa39f6f8dcd54644d55edd461d73d0 | 91d155ea92da372b319a79dd4eef69533ee15170 | [`CI update`] Try to use dockers and no cache (#29202)
* change cis
* nits
* update
* minor updates
* [push-ci-image]
* nit [push-ci-image]
* nitsssss
* [build-ci-image]
* [push-ci-image]
* [push-ci-image]
* both
* [push-ci-image]
* this?
* [push-ci-image]
* pypi-kenlm needs g... | [
{
"path": ".circleci/config.yml",
"patch": "@@ -12,7 +12,7 @@ jobs:\n # Ensure running with CircleCI/huggingface\n check_circleci_user:\n docker:\n- - image: cimg/python:3.8.12\n+ - image: python:3.10-slim\n parallelism: 1\n steps:\n - run: e... | 2024-05-06T08:10:32 |
nodejs/node | 556ebab30e171c2941deb529a42ad47301972af5 | 801e61ad5a4ee2e5621c2578a44c9d63c99c9ff6 | child_process: restore exec{File}Sync error props
In PR [1], a bunch of properties were removed from the error thrown by
execSync and execFileSync. It turns out that some of those were still
supposed to be there, as the documentation states that the error
contains the entire result from the spawnSync call.
[1] https:... | [
{
"path": "lib/child_process.js",
"patch": "@@ -574,8 +574,7 @@ function checkExecSyncError(ret, args, cmd) {\n err = new Error(msg);\n }\n if (err) {\n- err.status = ret.status < 0 ? errname(ret.status) : ret.status;\n- err.signal = ret.signal;\n+ Object.assign(err, ret);\n }\n return ... | 2017-10-07T04:06:18 |
ollama/ollama | e1a69d44c9c9adae94a54e123a7118915d6bbf4e | 3d620f9462b800b880d51e0eb6e51a91182f79db | Update faq.md (#1299)
Fix a typo in the CA update command | [
{
"path": "docs/faq.md",
"patch": "@@ -139,7 +139,7 @@ Ensure the certificate is installed as a system certificate when using HTTPS. Th\n ```dockerfile\n FROM ollama/ollama\n COPY my-ca.pem /usr/local/share/ca-certificates/my-ca.crt\n-RUN update-ca-certificate\n+RUN update-ca-certificates\n ```\n \n Build a... | 2023-11-28T14:54:42 |
ggml-org/llama.cpp | 81ab64f3c858c0db8c7c3a6bccd4cbbe624f52a3 | 8af1f5f430baaab1719db8f0a259bcc2a1cfdaa0 | ggml-cuda: enable cuda-graphs for `n-cpu-moe` (#18934)
* ggml-cuda: add split-wise cuda graph
* add n-cpu-moe compare_llama_bench.py
* fix hip/musa builds | [
{
"path": "ggml/src/ggml-cuda/common.cuh",
"patch": "@@ -1327,10 +1327,44 @@ struct ggml_backend_cuda_context {\n cudaStream_t streams[GGML_CUDA_MAX_DEVICES][GGML_CUDA_MAX_STREAMS] = { { nullptr } };\n cublasHandle_t cublas_handles[GGML_CUDA_MAX_DEVICES] = {nullptr};\n \n- std::unique_ptr<ggml_cu... | 2026-01-24T06:25:20 |
huggingface/transformers | deb7605a2a805eb0ef11f2706471c3d7cd68bfc0 | d0c72c15c25e100860b6af692d084f06546a0b7a | Prevent `TextGenerationPipeline._sanitize_parameters` from overriding previously provided parameters (#30362)
* Fixed TextGenerationPipeline._sanitize_parameters default params
* removed empty spaces
---------
Co-authored-by: Ng, Yen Ting <yen.ting.ng@intel.com> | [
{
"path": "src/transformers/pipelines/text_generation.py",
"patch": "@@ -129,19 +129,25 @@ def _sanitize_parameters(\n prefix=None,\n handle_long_generation=None,\n stop_sequence=None,\n- add_special_tokens=False,\n truncation=None,\n- padding=False,\n m... | 2024-05-03T15:49:28 |
nodejs/node | 801e61ad5a4ee2e5621c2578a44c9d63c99c9ff6 | 839faae45a915d6eda9681381ff001039b1afb49 | deps: cherry-pick 37a3a15c3 from V8 upstream
Original commit message:
[api] Intercept DefineProperty after Descriptor query
Analog to other interceptors, intercept the DefineProperty
call only after obtaining the property descriptor.
This behavior allows us to mirror calls on a sandboxed object
as it is ne... | [
{
"path": "common.gypi",
"patch": "@@ -27,7 +27,7 @@\n \n # Reset this number to 0 on major V8 upgrades.\n # Increment by one for each non-official patch applied to deps/v8.\n- 'v8_embedder_string': '-node.3',\n+ 'v8_embedder_string': '-node.4',\n \n # Enable disassembler for `--print-code... | 2017-10-18T14:30:34 |
ggml-org/llama.cpp | b5b8fa1c8b3b27683b2965a22f9985eec683d384 | a14b960bc70a0b48405409bbe3e0d6238473a0f8 | chat : fix translategemma crash on common_chat_format_example (#19019) | [
{
"path": "common/chat.cpp",
"patch": "@@ -2650,6 +2650,45 @@ static common_chat_params common_chat_params_init_exaone_moe(const common_chat_t\n return data;\n }\n \n+static common_chat_params common_chat_params_init_translate_gemma(const common_chat_template & tmpl, const struct templates_params & inpu... | 2026-01-23T11:03:42 |
golang/go | 8860251db862a091ec70e3571f0885d232187119 | 5853b4ee47f41f2d5ea8e63a185b2cfd6e08b98c | all: update vendored dependencies during Go 1.16 development
The Go 1.16 development cycle has started. This is the time to update
all golang.org/x/... module versions that contribute packages to the
std and cmd modules in the standard library to latest master versions.
Those versions have already gone through code r... | [
{
"path": "src/cmd/go.mod",
"patch": "@@ -1,14 +1,13 @@\n module cmd\n \n-go 1.15\n+go 1.16\n \n require (\n \tgithub.com/google/pprof v0.0.0-20200229191704-1ebb73c60ed3\n \tgithub.com/ianlancetaylor/demangle v0.0.0-20200414190113-039b1ae3a340 // indirect\n-\tgolang.org/x/arch v0.0.0-20200511175325-f7c78586... | 2020-09-17T16:32:24 |
ollama/ollama | e9216ea459f3fabaef81f376ddbf0ba6ef292b37 | 9e4a316405d9c5ea37ee21bba4f99573db276c70 | fix readline history on linux | [
{
"path": "readline/history.go",
"patch": "@@ -43,9 +43,12 @@ func (h *History) Init() error {\n \t}\n \n \tpath := filepath.Join(home, \".ollama\", \"history\")\n+\tif err := os.MkdirAll(filepath.Dir(path), 0o755); err != nil {\n+\t\treturn err\n+\t}\n+\n \th.Filename = path\n \n-\t//todo check if the file... | 2023-11-26T20:59:04 |
huggingface/transformers | 66f675eb656078a00469b223a65f88fd11790857 | 425e1a0426c8da23a00c456fd69745a1b3e6db5f | Fix W&B run name (#30462)
* Remove comparison to output_dir
* Update docs for `run_name`
* Add warning | [
{
"path": "src/transformers/integrations/integration_utils.py",
"patch": "@@ -762,9 +762,14 @@ def setup(self, args, state, model, **kwargs):\n if trial_name is not None:\n init_args[\"name\"] = trial_name\n init_args[\"group\"] = args.run_name\n- else:... | 2024-05-03T11:04:15 |
nodejs/node | 27e12e75244343ffcde4b899a31e88d452fa4d6e | b5569dbe8dee7f259644ef6b73b756d9ea2ba090 | tty,doc: add type-check to isatty
Previously, various inputs other than non-negative integers would
produce incorrect results.
Added type-checking on input, returning false for anything other than
non-negative integers.
Also clarified in docs.
PR-URL: https://github.com/nodejs/node/pull/15567
Reviewed-By: Luigi Pin... | [
{
"path": "doc/api/tty.md",
"patch": "@@ -123,4 +123,5 @@ added: v0.5.8\n * `fd` {number} A numeric file descriptor\n \n The `tty.isatty()` method returns `true` if the given `fd` is associated with\n-a TTY and `false` if is not.\n+a TTY and `false` if it is not, including whenever `fd` is not a non-negativ... | 2017-09-23T08:40:35 |
ggml-org/llama.cpp | 091a46cb8d43c0e662d04b80a3d11320d25b7d49 | a3e812811d8f12f4236efa41287dc3dcd5c3c2f6 | ggml-cpu: aarm64: q5_K repack gemm and gemv (and generic) implementations (i8mm) (#18860)
* Boilerplate for q5_Kx8 REPACK on ARM and fallback
Signed-off-by: Alberto Cabrera <alberto.cabrera@liquid.ai>
* Implements make_block_q5_Kx8 by extending make_block_q4_Kx8
Signed-off-by: Alberto Cabrera <alberto.cabrera@liqui... | [
{
"path": "ggml/src/ggml-cpu/arch-fallback.h",
"patch": "@@ -38,19 +38,21 @@\n #define ggml_gemv_q4_0_4x4_q8_0_generic ggml_gemv_q4_0_4x4_q8_0\n #define ggml_gemv_q4_0_4x8_q8_0_generic ggml_gemv_q4_0_4x8_q8_0\n #define ggml_gemv_q4_0_8x8_q8_0_generic ggml_gemv_q4_0_8x8_q8_0\n+#define ggml_gemv_q2_K_8x8_q8_K... | 2026-01-23T07:55:08 |
huggingface/transformers | 425e1a0426c8da23a00c456fd69745a1b3e6db5f | a0e77a1f6bdfcceccdc5618e8a01ee32ef47bfa8 | add mlp bias for llama models (#30031)
* add bias
* fix quality | [
{
"path": "src/transformers/models/cohere/modeling_cohere.py",
"patch": "@@ -161,7 +161,6 @@ def apply_rotary_pos_emb(q, k, cos, sin, position_ids=None, unsqueeze_dim=1):\n return q_embed.to(dtype=dtype), k_embed.to(dtype=dtype)\n \n \n-# Copied from transformers.models.llama.modeling_llama.LlamaMLP Lla... | 2024-05-03T09:02:17 |
golang/go | f92c64045f5effd4339749b8ce3b63b88cfef4d4 | 7e9369a517d9ebf867748719948d8cbccec3bc57 | debug/dwarf: speed up SkipChildren for compilation units
For a common pattern of iterating only over top-level compilation units (CU)
Reader.SkipChildren has decode and meterialize all CU subentries just
to skip them, because DW_TAG_compile_unit does not have DW_AT_sibling.
However, CUs have total size encoded before ... | [
{
"path": "src/debug/dwarf/entry.go",
"patch": "@@ -717,6 +717,7 @@ type Reader struct {\n \td *Data\n \terr error\n \tunit int\n+\tlastUnit bool // set if last entry returned by Next is TagCompileUnit/TagPartialUnit\n \tlastChildren bool // .Children of last entry return... | 2020-09-21T10:19:47 |
nodejs/node | b5569dbe8dee7f259644ef6b73b756d9ea2ba090 | 02a52670b832b17b90365aa42b5620c20fe357f6 | test: handle blank shells in test-os.js
The shell in /etc/passwd can be blank, in which case the user is given
the default shell. Handle this by only checking the shell contains a
path separator if the string isn't empty.
PR-URL: https://github.com/nodejs/node/pull/16287
Fixes: https://github.com/nodejs/node/issues/1... | [
{
"path": "test/parallel/test-os.js",
"patch": "@@ -208,7 +208,11 @@ if (common.isWindows) {\n } else {\n is.number(pwd.uid);\n is.number(pwd.gid);\n- assert.ok(pwd.shell.includes(path.sep));\n+ assert.strictEqual(typeof pwd.shell, 'string');\n+ // It's possible for /etc/passwd to leave the user's sh... | 2017-10-18T09:02:58 |
ggml-org/llama.cpp | a5eaa1d6a3732bc0f460b02b61c95680bba5a012 | e2baf02162382a14c9f4fc15d7681a715256453c | mla : make the V tensor a view of K (#18986)
* mla : pass V as a view of K to the FA op
* cuda : adjust mla logic to new layout
* kv-cache : fix rope shift
* tests : remove comment
* cuda : fix reusable_cutoff
Co-authored-by: Johannes Gäßler <johannesg@5d6.de>
---------
Co-authored-by: Johannes Gäßler <johannes... | [
{
"path": "ggml/src/ggml-cuda/fattn-common.cuh",
"patch": "@@ -778,12 +778,15 @@ void launch_fattn(\n ) {\n constexpr int ncols = ncols1 * ncols2;\n \n- const bool is_mla = DV == 512; // TODO better parameterization\n-\n const ggml_tensor * Q = dst->src[0];\n const ggml_tensor * K = dst->src[... | 2026-01-22T20:09:01 |
golang/go | 7e9369a517d9ebf867748719948d8cbccec3bc57 | eda1d405444262ab36160b87179e086872804709 | cmd/link: add go.mod to TestFuncAlign
Fixes #41531
Change-Id: I8b4f0d5b7094e56787998d244d8a4c03becb8452
Reviewed-on: https://go-review.googlesource.com/c/go/+/256302
Trust: Jay Conrod <jayconrod@google.com>
Reviewed-by: Bryan C. Mills <bcmills@google.com> | [
{
"path": "src/cmd/link/link_test.go",
"patch": "@@ -595,7 +595,12 @@ func TestFuncAlign(t *testing.T) {\n \t}\n \tdefer os.RemoveAll(tmpdir)\n \n-\tsrc := filepath.Join(tmpdir, \"falign.go\")\n+\tsrc := filepath.Join(tmpdir, \"go.mod\")\n+\terr = ioutil.WriteFile(src, []byte(\"module cmd/link/TestFuncAlign... | 2020-09-21T16:10:16 |
nodejs/node | 838eca26566134af60116ebc9694e860d7310f0d | 7a171fd88a2e7ce06121ce016ea0f15d1b5a158e | src: operator[] checks bounds in debug mode
PR-URL: https://github.com/nodejs/node/pull/16002
Reviewed-By: Ruben Bridgewater <ruben@bridgewater.de>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Timothy Gu <timothygu99@gmail.com>
Reviewed-By: Anna Henningsen <anna@addaleax.net> | [
{
"path": "src/string_search.h",
"patch": "@@ -44,7 +44,9 @@ class Vector {\n \n // Access individual vector elements - checks bounds in debug mode.\n T& operator[](size_t index) const {\n+#ifdef DEBUG\n CHECK(index < length_);\n+#endif\n return start_[is_forward_ ? index : (length_ - index - 1)... | 2017-10-06T18:24:47 |
ggml-org/llama.cpp | e2baf02162382a14c9f4fc15d7681a715256453c | e34d6d03b25d9e8d07f3bd0190b27d0d01a7e416 | CUDA: fix alignment check for FA (#19023) | [
{
"path": "ggml/src/ggml-cuda/fattn.cu",
"patch": "@@ -46,7 +46,7 @@ static void ggml_cuda_flash_attn_ext_mma_f16_switch_ncols2(ggml_backend_cuda_con\n // are put into the template specialization without GQA optimizations.\n bool use_gqa_opt = mask && max_bias == 0.0f && K->ne[1] % FATTN_KQ_STRI... | 2026-01-22T19:39:25 |
huggingface/transformers | a0e77a1f6bdfcceccdc5618e8a01ee32ef47bfa8 | 59952994c422ae9fc17f8a3134c00396c948c42e | Fix CI after #30410 (#30612)
* Fix CI after #30410
* [run-slow] blenderbot | [
{
"path": "src/transformers/models/blenderbot/modeling_blenderbot.py",
"patch": "@@ -1067,8 +1067,8 @@ def __init__(self, config: BlenderbotConfig):\n super().__init__(config)\n \n padding_idx, vocab_size = config.pad_token_id, config.vocab_size\n- self.shared = nn.Embedding(vocab_siz... | 2024-05-02T20:18:48 |
ollama/ollama | 9fb5e8399c61217a4deef60a773de19ebd87caca | 82b9b329ff84506445f85aed8d65690bd840831a | Fix issues with inputting and formatting multi line strings in `ollama run`
Co-authored-by: Wen Sun <iwendellsun@gmail.com> | [
{
"path": "cmd/cmd.go",
"patch": "@@ -602,22 +602,20 @@ func generateInteractive(cmd *cobra.Command, model string, wordWrap bool, format\n \t\tfmt.Fprintln(os.Stderr, \"\")\n \t}\n \n-\tprompt := readline.Prompt{\n+\tscanner, err := readline.New(readline.Prompt{\n \t\tPrompt: \">>> \",\n \t\tAltProm... | 2023-11-26T04:30:34 |
golang/go | eda1d405444262ab36160b87179e086872804709 | 02ff8b8ce49c7c8f1180ee6e915b867368ad77d3 | make.bash: avoid warning when bootstrap doesn't support GOOS/GOARCH
Currently, if make.bash is run with a GOOS or GOARCH that the
bootstrap toolchain doesn't support, it will print an ominous but
harmless warning like:
2020/09/21 09:05:27 unsupported GOARCH arm64
This comes from the invocation of "go version" to g... | [
{
"path": "src/make.bash",
"patch": "@@ -167,7 +167,10 @@ if [ ! -x \"$GOROOT_BOOTSTRAP/bin/go\" ]; then\n \techo \"Set \\$GOROOT_BOOTSTRAP to a working Go tree >= Go 1.4.\" >&2\n \texit 1\n fi\n-GOROOT_BOOTSTRAP_VERSION=$($GOROOT_BOOTSTRAP/bin/go version | sed 's/go version //')\n+# Get the exact bootstrap... | 2020-09-21T13:18:32 |
ggml-org/llama.cpp | 94242a62c055700aac8a573853cbf7b45849975e | 6b99a223e35a24d68a6875b96829c945768c296d | ggml-zdnn : mark zDNN buffers as non-host (#18967)
While buffers reside in host memory,
additional transformation is needed to use buffers with zDNN.
Fixes #18848 | [
{
"path": "ggml/src/ggml-zdnn/ggml-zdnn.cpp",
"patch": "@@ -372,7 +372,8 @@ static size_t ggml_backend_zdnn_buffer_type_get_alignment(ggml_backend_buffer_ty\n }\n \n static bool ggml_backend_zdnn_buffer_type_is_host(ggml_backend_buffer_type_t buft) {\n- return true;\n+ /* while it resides in host memo... | 2026-01-22T00:16:21 |
rust-lang/rust | 38b364b29d67e8518614d19e18d822786f94d220 | 617aad8c2e8783f6df8e5d1f8bb1e4bcdc70aa7b | rustc_fluent_macro: use CARGO_CRATE_NAME instead of CARGO_PKG_NAME
fixes #137815 | [
{
"path": "compiler/rustc_fluent_macro/src/fluent.rs",
"patch": "@@ -78,8 +78,8 @@ fn failed(crate_name: &Ident) -> proc_macro::TokenStream {\n \n /// See [rustc_fluent_macro::fluent_messages].\n pub(crate) fn fluent_messages(input: proc_macro::TokenStream) -> proc_macro::TokenStream {\n- let crate_name ... | 2025-03-01T03:14:06 |
nodejs/node | 7a171fd88a2e7ce06121ce016ea0f15d1b5a158e | 01a55ee8cb2798217c8507fac8ac65249c105d54 | http2: small fixes to http2 core
A few small changes:
- fix debug statement location
- simplify conditional with returns
- consistent use of template strings
PR-URL: https://github.com/nodejs/node/pull/16327
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Luigi Pinca <luigipinca@gmail.com> | [
{
"path": "lib/internal/http2/core.js",
"patch": "@@ -881,14 +881,14 @@ class Http2Session extends EventEmitter {\n \n // A stream cannot be made to depend on itself\n if (options.parent === id) {\n- debug(`[${sessionName(this[kType])}] session still connecting. queue ` +\n- 'priorit... | 2017-10-18T23:07:41 |
huggingface/transformers | 59952994c422ae9fc17f8a3134c00396c948c42e | 4c940934dad4cad6992d8470a71942db6ab6c0ac | Add HQQ quantization support (#29637)
* update HQQ transformers integration
* push import_utils.py
* add force_hooks check in modeling_utils.py
* fix | with Optional
* force bias as param
* check bias is Tensor
* force forward for multi-gpu
* review fixes pass
* remove torch grad()
* if any ... | [
{
"path": "docker/transformers-quantization-latest-gpu/Dockerfile",
"patch": "@@ -45,6 +45,9 @@ RUN python3 -m pip install --no-cache-dir git+https://github.com/huggingface/opt\n # Add aqlm for quantization testing\n RUN python3 -m pip install --no-cache-dir aqlm[gpu]==1.0.2\n \n+# Add hqq for quantization ... | 2024-05-02T16:51:49 |
ollama/ollama | 12e8c12d2b5c0658cad014b58a8baf597b0741df | d77dde126b5fc6e340a9e65f1b9e33316a2c760c | Disable CUDA peer access as a workaround for multi-gpu inference bug (#1261)
When CUDA peer access is enabled, multi-gpu inference will produce
garbage output. This is a known bug of llama.cpp (or nvidia). Until the
upstream bug is fixed, we can disable CUDA peer access temporarily
to ensure correct output.
See ... | [
{
"path": "llm/llama.cpp/generate_linux.go",
"patch": "@@ -21,6 +21,6 @@ package llm\n //go:generate cmake -S ggml -B ggml/build/cuda -DLLAMA_CUBLAS=on -DLLAMA_ACCELERATE=on -DLLAMA_K_QUANTS=on\n //go:generate cmake --build ggml/build/cuda --target server --config Release\n //go:generate mv ggml/build/cuda/... | 2023-11-24T19:05:57 |
golang/go | eb3c6a93c3236bbde5dee6cc5bd4ca9f8ab1647a | b4ea67200977b99ede1885ed77e034a2fdf434f5 | runtime: disable stack shrinking in activeStackChans race window
Currently activeStackChans is set before a goroutine blocks on a channel
operation in an unlockf passed to gopark. The trouble is that the
unlockf is called *after* the G's status is changed, and the G's status
is what is used by a concurrent mark worker... | [
{
"path": "src/runtime/chan.go",
"patch": "@@ -250,6 +250,11 @@ func chansend(c *hchan, ep unsafe.Pointer, block bool, callerpc uintptr) bool {\n \tgp.waiting = mysg\n \tgp.param = nil\n \tc.sendq.enqueue(mysg)\n+\t// Signal to anyone trying to shrink our stack that we're about\n+\t// to park on a channel. ... | 2020-08-10T20:02:22 |
rust-lang/rust | 6d07144613ffb54e6856bd5543acfe5e12eb3447 | b1196717fcba13d8ff73ec06c3d8c5bee3b74ec9 | Fix logic error in Buffer::read_more()
Buffer::read_more() is supposed to refill the buffer without discarding
its contents, which are in the range `pos .. filled`.
It mistakenly borrows the range `pos ..`, fills that, and then
increments `filled` by the amount read. This overwrites the buffer's
existing contents and... | [
{
"path": "library/std/src/io/buffered/bufreader/buffer.rs",
"patch": "@@ -109,8 +109,8 @@ impl Buffer {\n \n /// Read more bytes into the buffer without discarding any of its contents\n pub fn read_more(&mut self, mut reader: impl Read) -> io::Result<usize> {\n- let mut buf = BorrowedBuf::fr... | 2025-03-01T01:36:19 |
ggml-org/llama.cpp | 3802d3c78f5791f77ee9706e8ef440ca61d12ea3 | 9da3dcd75313b4c492b1b3b0e189ecc70906a195 | fix: Use `tabular-nums` for chat message statistics (#18915)
* fix: Use `tabular-nums` for chat message statistics
* fix: Rebuild WebUI | [
{
"path": "tools/server/webui/src/lib/components/app/chat/ChatMessages/ChatMessageAssistant.svelte",
"patch": "@@ -249,7 +249,7 @@\n \t\t</div>\n \t{/if}\n \n-\t<div class=\"info my-6 grid gap-4\">\n+\t<div class=\"info my-6 grid gap-4 tabular-nums\">\n \t\t{#if displayedModel()}\n \t\t\t<div class=\"inline... | 2026-01-21T17:46:01 |
huggingface/transformers | 39359e5b5f6246c667cc25f743437cb05d492ea7 | 9719202d37492a43c323d03a81aed23d14d98dec | Fix FX tracing issues for Llama (#30619) | [
{
"path": "src/transformers/utils/fx.py",
"patch": "@@ -714,9 +714,14 @@ class HFCacheProxy(HFProxy):\n Proxy that represents an instance of `transformers.cache_utils.Cache`.\n \"\"\"\n \n+ def install_orig_cache_cls(self, orig_cache_cls: Type[Cache]):\n+ self._orig_cache_cls = orig_cache_... | 2024-05-02T15:03:10 |
nodejs/node | 023dc4bb83d9a8cacc9d77a91f21b719dfdce756 | 58d8fc4c10f10bdec747529ad4404dcbdbed7eb7 | http2: cleanup access via Symbols in core
Minimize property access via Symbols by consistently creating
new constants for things that are retrieved multiple times
within a method or a function. Fix cases where existing
constants aren't used where appropriate.
PR-URL: https://github.com/nodejs/node/pull/16327
Reviewed... | [
{
"path": "lib/internal/http2/core.js",
"patch": "@@ -145,8 +145,9 @@ function emit(self, ...args) {\n // the block of headers on.\n function onSessionHeaders(id, cat, flags, headers) {\n const owner = this[kOwner];\n+ const type = owner[kType];\n _unrefActive(owner);\n- debug(`[${sessionName(owner[kT... | 2017-10-18T16:02:19 |
golang/go | b4ea67200977b99ede1885ed77e034a2fdf434f5 | 331614c4daa5504ddfe35a96371cc34783d14cf1 | cmd/go: default to GO111MODULE=on
Fixes #41330
Change-Id: Ib66087ac5e1eb827694915f21c32c4d88e1b9a7e
Reviewed-on: https://go-review.googlesource.com/c/go/+/255052
Run-TryBot: Jay Conrod <jayconrod@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Bryan C. Mills <bcmills@google.com>
Trust: Jay Conrod <j... | [
{
"path": "src/cmd/go/internal/modload/init.go",
"patch": "@@ -120,9 +120,9 @@ func Init() {\n \tswitch env {\n \tdefault:\n \t\tbase.Fatalf(\"go: unknown environment setting GO111MODULE=%s\", env)\n-\tcase \"auto\", \"\":\n+\tcase \"auto\":\n \t\tmustUseModules = ForceUseModules\n-\tcase \"on\":\n+\tcase \... | 2020-09-15T16:59:05 |
rust-lang/rust | b1196717fcba13d8ff73ec06c3d8c5bee3b74ec9 | 60493b8973ac5ba632952eaa2f212b56bb97ccfe | Tweak BufReader::peek() doctest to expose bug in Buffer::read_more()
This patch makes BufReader::peek()'s doctest call read_more() to refill
the buffer before the inner reader hits EOF. This exposes a bug in
read_more() that causes an out-of-bounds slice access and segfault. | [
{
"path": "library/std/src/io/buffered/bufreader.rs",
"patch": "@@ -118,16 +118,16 @@ impl<R: Read + ?Sized> BufReader<R> {\n /// #![feature(bufreader_peek)]\n /// use std::io::{Read, BufReader};\n ///\n- /// let mut bytes = &b\"oh, hello\"[..];\n+ /// let mut bytes = &b\"oh, hello there\"... | 2025-03-01T01:30:53 |
ggml-org/llama.cpp | bd544c94a34366b009901b47d4850bf9a9bce4d1 | 14be5a39b1286fc514efd19119c8f1c2e545634f | vulkan: Remove transfer_ctx, do everything in compute_ctx. (#18945)
* vulkan: Remove transfer_ctx, do everything in compute_ctx.
We had a bug where a set_tensor_async (using transfer_ctx) didn't get
submitted before the graph_compute (using compute_ctx) that came after
it. To avoid this sort of issue, just do everyth... | [
{
"path": "ggml/src/ggml-vulkan/ggml-vulkan.cpp",
"patch": "@@ -1813,7 +1813,6 @@ struct ggml_backend_vk_context {\n bool prealloc_x_need_sync, prealloc_y_need_sync, prealloc_split_k_need_sync;\n \n vk_context_ref compute_ctx;\n- vk_context_ref transfer_ctx;\n \n std::vector<vk_context_ref> t... | 2026-01-21T17:01:40 |
huggingface/transformers | 9719202d37492a43c323d03a81aed23d14d98dec | 66abe1395157f8cb18830166625d886671eeb2fb | Generate: fix `SinkCache` on Llama models (#30581) | [
{
"path": "src/transformers/cache_utils.py",
"patch": "@@ -207,7 +207,9 @@ def __init__(self, window_length: int, num_sink_tokens: int) -> None:\n self.value_cache: List[torch.Tensor] = []\n self.window_length = window_length\n self.num_sink_tokens = num_sink_tokens\n- self.co... | 2024-05-02T14:24:33 |
nodejs/node | 53f4ac2774fa8aa1e964cf5fb660bcc72433e0b9 | fe8297ca92874770880e09e052c94f6ece5b2939 | test: increase enoughTestMem to 1.75 Gb
PR-URL: https://github.com/nodejs/node/pull/16374
Fixes: https://github.com/nodejs/node/issues/16354
Reviewed-By: Gireesh Punathil <gpunathi@in.ibm.com>
Reviewed-By: Joyee Cheung <joyeec9h3@gmail.com>
Reviewed-By: Luigi Pinca <luigipinca@gmail.com>
Reviewed-By: Gibson Fahnestock... | [
{
"path": "test/common/index.js",
"patch": "@@ -53,7 +53,7 @@ exports.isFreeBSD = process.platform === 'freebsd';\n exports.isLinux = process.platform === 'linux';\n exports.isOSX = process.platform === 'darwin';\n \n-exports.enoughTestMem = os.totalmem() > 0x40000000; /* 1 Gb */\n+exports.enoughTestMem = o... | 2017-10-22T02:40:06 |
golang/go | 331614c4daa5504ddfe35a96371cc34783d14cf1 | a3868028ac8470d1ab7782614707bb90925e7fe3 | runtime: improve error messages after allocating a stack that is too big
In the current implementation, we can observe crashes after calling
debug.SetMaxStack and allocating a stack larger than 4GB since
stackalloc works with 32-bit sizes. To avoid this, we define an upper
limit as the largest feasible point we can gr... | [
{
"path": "src/runtime/debug/garbage.go",
"patch": "@@ -106,6 +106,8 @@ func FreeOSMemory() {\n // the program crashes.\n // SetMaxStack returns the previous setting.\n // The initial setting is 1 GB on 64-bit systems, 250 MB on 32-bit systems.\n+// There may be a system-imposed maximum stack limit regardle... | 2020-09-18T16:15:41 |
ggml-org/llama.cpp | 14be5a39b1286fc514efd19119c8f1c2e545634f | fbbf3ad1900bbaa97cd3c8de4c764afb0f6d8972 | common : improve error message when HTTPS is missing but required (#18987)
Signed-off-by: Adrien Gallouët <angt@huggingface.co> | [
{
"path": "common/http.h",
"patch": "@@ -57,6 +57,17 @@ static std::pair<httplib::Client, common_http_url> common_http_client(const std:\n throw std::runtime_error(\"error: invalid URL format\");\n }\n \n+#ifndef CPPHTTPLIB_OPENSSL_SUPPORT\n+ if (parts.scheme == \"https\") {\n+ throw s... | 2026-01-21T16:58:38 |
ollama/ollama | 199941cd151fb789577af77776fc970c6db99a3a | 37d95157dfed0726d751d3a694e9d6cd3c923097 | fix: gguf int type | [
{
"path": "llm/gguf.go",
"patch": "@@ -335,7 +335,7 @@ func (llm *ggufModel) readArrayV1(r io.Reader) (arr []any, err error) {\n \t\tcase ggufTypeUint8:\n \t\t\tarr = append(arr, llm.readU8(r))\n \t\tcase ggufTypeInt8:\n-\t\t\tarr = append(arr, llm.readU8(r))\n+\t\t\tarr = append(arr, llm.readI8(r))\n \t\tc... | 2023-11-22T19:40:30 |
huggingface/transformers | aa55ff44a259913f0a5254a542b1d7af4b85e037 | 801894e08c53a80c2b4377ce102a44dccfda8928 | Docs: fix `generate`-related rendering issues (#30600)
* does this work?
* like this?
* fix the other generate links
* missing these | [
{
"path": "docs/source/en/generation_strategies.md",
"patch": "@@ -21,7 +21,7 @@ more. It also plays a role in a variety of mixed-modality applications that have\n and vision-to-text. Some of the models that can generate text include\n GPT2, XLNet, OpenAI GPT, CTRL, TransformerXL, XLM, Bart, T5, GIT, Whispe... | 2024-05-02T13:42:25 |
nodejs/node | cfea677a5ee82db7c031adf0cdd828e56a5ccd0b | 438e7fdaf2f0c05e964bc988ff2b6e3d273de1cb | test: use fixtures.readKey in https-timeout-server
Replace `common.fixturesDir` usage in test-https-timeout-server.js with
`fixtures.readKey()`.
PR-URL: https://github.com/nodejs/node/pull/15871
Reviewed-By: Joyee Cheung <joyeec9h3@gmail.com>
Reviewed-By: Gireesh Punathil <gpunathi@in.ibm.com> | [
{
"path": "test/parallel/test-https-timeout-server.js",
"patch": "@@ -25,15 +25,16 @@ const common = require('../common');\n if (!common.hasCrypto)\n common.skip('missing crypto');\n \n+const fixtures = require('../common/fixtures');\n+\n const assert = require('assert');\n const https = require('https');... | 2017-10-06T17:01:58 |
golang/go | 4c4a376736fff47b08ab6053605c3b68d87552b5 | 8925290cf701fc8f7ec95e4df3d6a8d423b26780 | runtime/debug: skip fault address test on unsupported platforms
Change-Id: I6a6fe616365fa542218fbc9ff61805ff70a1ef63
Reviewed-on: https://go-review.googlesource.com/c/go/+/255999
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Reviewed-by: Bryan C. Mills <bcmills@google.com>
Trust:... | [
{
"path": "src/runtime/debug/panic_test.go",
"patch": "@@ -9,13 +9,20 @@\n package debug_test\n \n import (\n+\t\"runtime\"\n \t\"runtime/debug\"\n \t\"syscall\"\n \t\"testing\"\n \t\"unsafe\"\n )\n \n func TestPanicOnFault(t *testing.T) {\n+\tif runtime.GOARCH == \"s390x\" {\n+\t\tt.Skip(\"s390x fault addr... | 2020-09-18T20:01:54 |
huggingface/transformers | 801894e08c53a80c2b4377ce102a44dccfda8928 | f57f014936404f32818cfec0ce977fb805bf259c | phi3 chat_template does not support system role (#30606)
* phi3 chat_template does not support system role
* fix doc test error | [
{
"path": "docs/source/en/model_doc/phi3.md",
"patch": "@@ -9,7 +9,7 @@ Unless required by applicable law or agreed to in writing, software distributed\n an \"AS IS\" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the\n specific language governing permiss... | 2024-05-02T13:30:21 |
ggml-org/llama.cpp | fbbf3ad1900bbaa97cd3c8de4c764afb0f6d8972 | 33f890e5799d1bb539f72f3d7e1ea30b8cff1123 | server: /v1/responses (partial) (#18486)
* from previous PR
* Make instruction(system) as first message
* Convert [input_message] (text/image/file)
* Rename convert_responses_to_chatcmpl(body) -> response_body
* Initial tool call support
* Erase instructions field from chatcmpl body
* Feed reasoning texts to cha... | [
{
"path": "requirements/requirements-tool_bench.txt",
"patch": "@@ -3,7 +3,7 @@ pytest~=8.3.3\n huggingface_hub>=0.34.0,<1.0\n matplotlib~=3.10.0\n numpy~=1.26.4\n-openai~=1.55.3\n+openai~=2.14.0\n pandas~=2.2.3\n prometheus-client~=0.20.0\n requests~=2.32.3",
"additions": 1,
"deletions": 1,
"la... | 2026-01-21T16:47:23 |
ollama/ollama | 37d95157dfed0726d751d3a694e9d6cd3c923097 | 2eaa95b417e0711f90fe298ef233e58c6fe0a478 | fix relative path on create (#1222) | [
{
"path": "server/images.go",
"patch": "@@ -228,27 +228,32 @@ func GetModel(name string) (*Model, error) {\n \treturn model, nil\n }\n \n-func realpath(p string) string {\n-\tabspath, err := filepath.Abs(p)\n+func realpath(mfDir, from string) string {\n+\tabspath, err := filepath.Abs(from)\n \tif err != nil... | 2023-11-21T20:43:17 |
nodejs/node | 8e268c70dd5b5d98658544a5d46cd8d8d4b17481 | fe4675b301f0e3e26dcb88ef9f68fe35403330a0 | test: fix common.PIPE path bug
`common.PIPE` is returning a path name in `test` rather than in the
`tmp` directory for each test. This is causing multiple test failures in
CI. Make the path name inside the temporary directories again. This way
the pipe is removed by `common.refreshTmpDir()` on POSIX.
The bug in `comm... | [
{
"path": "test/common/index.js",
"patch": "@@ -274,7 +274,7 @@ Object.defineProperty(exports, 'hasFipsCrypto', {\n const localRelative = path.relative(process.cwd(), `${exports.tmpDir}/`);\n const pipePrefix = exports.isWindows ? '\\\\\\\\.\\\\pipe\\\\' : localRelative;\n const pipeName = `node-test.... | 2017-10-21T18:10:42 |
rust-lang/rust | 870f3952ac57f97cbb5f6d8cbd2d01e6ae3db7c5 | a6a78e13e2b79d9f93e7a2d1a914ff9a6c0cb3ba | Fix link to ty::Ty in clippy_utils | [
{
"path": "clippy_utils/src/ty/mod.rs",
"patch": "@@ -37,7 +37,7 @@ use crate::{def_path_def_ids, match_def_path, path_res};\n mod type_certainty;\n pub use type_certainty::expr_type_is_certain;\n \n-/// Lower a [`hir::Ty`] to a [`rustc_middle::Ty`].\n+/// Lower a [`hir::Ty`] to a [`rustc_middle::ty::Ty`].\... | 2025-02-21T16:42:48 |
huggingface/transformers | f57f014936404f32818cfec0ce977fb805bf259c | a65da83d75686fed9b0478bd361bc9e17184771c | Use `contiguous()` in clip checkpoint conversion script (#30613)
* fix
* fix
---------
Co-authored-by: ydshieh <ydshieh@users.noreply.github.com> | [
{
"path": "src/transformers/models/clip/convert_clip_original_pytorch_to_hf.py",
"patch": "@@ -82,15 +82,15 @@ def copy_encoder(hf_encoder, pt_model):\n \n def copy_text_model_and_projection(hf_model, pt_model):\n # copy projection\n- hf_model.text_projection.weight.data = pt_model.text_projection.da... | 2024-05-02T11:59:40 |
ggml-org/llama.cpp | 067b8d7af3f4775b1ad39470d1860696a8a7b3a3 | 50b7f076a55670645525955f54695f1c557c6f6a | Revert "vulkan: force full subgroups for flash attention to fix intel subgroup crash (#17356)" (#18831)
This reverts commit 980b7cd17e055c8c587f79ffda7eb4fddf405566. | [
{
"path": "ggml/src/ggml-vulkan/ggml-vulkan.cpp",
"patch": "@@ -3180,15 +3180,15 @@ static void ggml_vk_load_shaders(vk_device& device) {\n if (path == FAPATH) { \\\n if (aligned) { \\\n if (f32acc) { \\\n- ggml_vk_create_pipeline(device... | 2026-01-21T16:13:43 |
golang/go | 8925290cf701fc8f7ec95e4df3d6a8d423b26780 | 7e54aa2c25690f5a7f5baad112d231b6ff8d4e5e | reflect: use zero buffer to back the Value returned by Zero
In the common case (<1KB types), no allocation is required
by reflect.Zero.
Also use memclr instead of memmove in Set when the source
is known to be zero.
Fixes #33136
Change-Id: Ic66871930fbb53328032e587153ebd12995ccf55
Reviewed-on: https://go-review.goog... | [
{
"path": "src/reflect/all_test.go",
"patch": "@@ -6006,6 +6006,14 @@ func TestReflectMethodTraceback(t *testing.T) {\n \t}\n }\n \n+func TestSmallZero(t *testing.T) {\n+\ttype T [10]byte\n+\ttyp := TypeOf(T{})\n+\tif allocs := testing.AllocsPerRun(100, func() { Zero(typ) }); allocs > 0 {\n+\t\tt.Errorf(\"C... | 2020-09-18T21:19:22 |
ollama/ollama | f2113c1fc79110c5f243a39cb5ac03590b67fed9 | 6452e2ecb83ae03b1c504e09b6ca3ce9868d0c45 | fix potential error in progress bar calculation | [
{
"path": "progress/bar.go",
"patch": "@@ -95,7 +95,10 @@ func (b *Bar) String() string {\n \t}\n \n \t// 44 is the maximum width for the stats on the right of the progress bar\n-\tsuf.WriteString(strings.Repeat(\" \", 44-suf.Len()-len(timing)))\n+\tpad := 44 - suf.Len() - len(timing)\n+\tif pad > 0 {\n+\t\... | 2023-11-21T17:48:20 |
nodejs/node | fe4675b301f0e3e26dcb88ef9f68fe35403330a0 | c4c63812822fc973cb3d4bb4040729213dcaa620 | module: fix main resolution and not found updates
This simplifies the top-level load when ES modules are enabled
as we can entirely delegate the module resolver, which will hand
over to CommonJS where appropriate.
All not found errors are made consistent to throw during resolve
and have the MODULE_NOT_FOUND code.
In... | [
{
"path": "lib/internal/loader/Loader.js",
"patch": "@@ -47,6 +47,7 @@ class Loader {\n throw new errors.TypeError('ERR_INVALID_ARG_TYPE',\n 'parentURL', 'string');\n }\n+\n const { url, format } = await this.resolver(specifier, parentURL,\n ... | 2017-10-11T13:45:21 |
huggingface/transformers | a65da83d75686fed9b0478bd361bc9e17184771c | 4ad5adaf1d224fa28ffa8e1d124846b1d55a5d0e | fix:missing `output_router_logits` in SwitchTransformers (#30573)
* fix:missing `output_router_logits` in SwitchTransformers
* fix whitespace in blank line | [
{
"path": "src/transformers/models/switch_transformers/modeling_switch_transformers.py",
"patch": "@@ -1721,6 +1721,8 @@ def prepare_inputs_for_generation(\n \n input_ids = input_ids[:, remove_prefix_length:]\n \n+ output_router_logits = kwargs.get(\"output_router_logits\", True)\n+\n ... | 2024-05-02T11:47:00 |
rust-lang/rust | 1e4bce2ee1f815a1c4a7250730cd9e3b732061f2 | f5851e704531d3941e84afbd998a506ffe37a3d5 | Fix link to ty::Ty in clippy_utils | [
{
"path": "src/tools/clippy/clippy_utils/src/ty/mod.rs",
"patch": "@@ -37,7 +37,7 @@ use crate::{def_path_def_ids, match_def_path, path_res};\n mod type_certainty;\n pub use type_certainty::expr_type_is_certain;\n \n-/// Lower a [`hir::Ty`] to a [`rustc_middle::Ty`].\n+/// Lower a [`hir::Ty`] to a [`rustc_m... | 2025-02-21T16:42:48 |
golang/go | d91d0762c7757c12c7d5f9e2ae3f170d5bd7ba84 | 65dfe4a772a4bc612219d93886e5c07290785ee6 | runtime/debug: provide Addr method for errors from SetPanicOnFault
When we're building a panic that's triggered by a memory fault when
SetPanicOnFault has been called, include an Addr method. This
method reports the address at which the fault occurred.
Fixes #37023
RELNOTE=yes
Change-Id: Idff144587d6b75070fdc861a36... | [
{
"path": "src/runtime/debug/garbage.go",
"patch": "@@ -139,6 +139,11 @@ func SetMaxThreads(threads int) int {\n // manipulation of memory may cause faults at non-nil addresses in less\n // dramatic situations; SetPanicOnFault allows such programs to request\n // that the runtime trigger only a panic, not a... | 2020-08-20T21:22:30 |
ollama/ollama | 6452e2ecb83ae03b1c504e09b6ca3ce9868d0c45 | aabd71aede99443d5858213eec4e8d4c10713c55 | fix cases where progress bar would not be fixed size | [
{
"path": "progress/bar.go",
"patch": "@@ -89,17 +89,13 @@ func (b *Bar) String() string {\n \n \tfmt.Fprintf(&suf, \")\")\n \n-\telapsed := time.Since(b.started)\n \tvar timing string\n \tif stats.value > b.initialValue && stats.value < b.maxValue {\n-\t\ttiming = fmt.Sprintf(\"[%s:%s]\", formatDuration(el... | 2023-11-21T17:07:25 |
nodejs/node | c4c63812822fc973cb3d4bb4040729213dcaa620 | a627c5fc136b3233959e8e1bff90ae4b2d0ec322 | test: use fixtures.readKey instead of fixturesDir
PR-URL: https://github.com/nodejs/node/pull/15976
Reviewed-By: Ruben Bridgewater <ruben@bridgewater.de>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Tobias Nießen <tniessen@tnie.de>
Reviewed-By: Gireesh Punathil <gpunathi@in.ibm.com> | [
{
"path": "test/parallel/test-tls-delayed-attach.js",
"patch": "@@ -24,17 +24,17 @@ const common = require('../common');\n if (!common.hasCrypto)\n common.skip('missing crypto');\n \n+const fixtures = require('../common/fixtures');\n const assert = require('assert');\n const tls = require('tls');\n-const ... | 2017-10-06T17:55:49 |
ggml-org/llama.cpp | 12a4a47e6aaf691492644c39da453745aaee1672 | 37c35f0e1c625831687b146cbb0a57654ef88ca2 | Fix GLM 4.7 Lite MoE gating func (#18980)
* Fix GLM 4.7 MoE gating func
* Update src/models/deepseek2.cpp
Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com>
* Update src/llama-model.cpp
Co-authored-by: Xuan-Son Nguyen <thichthat@gmail.com>
---------
Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret... | [
{
"path": "src/llama-model.cpp",
"patch": "@@ -1713,7 +1713,12 @@ void llama_model::load_hparams(llama_model_loader & ml) {\n if (hparams.expert_gating_func == LLAMA_EXPERT_GATING_FUNC_TYPE_NONE) {\n // for compatibility with existing DeepSeek V2 and V2.5 GGUFs\n ... | 2026-01-21T11:35:20 |
huggingface/transformers | 4ad5adaf1d224fa28ffa8e1d124846b1d55a5d0e | f95302584b21db464423e0151dd6ce99ca07ccc3 | Fix copies for DBRX - neuron fix (#30610) | [
{
"path": "src/transformers/models/dbrx/modeling_dbrx.py",
"patch": "@@ -1256,8 +1256,11 @@ def _update_causal_mask(\n causal_mask = causal_mask.clone() # copy to contiguous memory for in-place edit\n if attention_mask.dim() == 2:\n mask_length = attention_mask.shape... | 2024-05-02T10:00:26 |
rust-lang/rust | 4f0e507165ae6ebf3304041324c196a6e5ae801c | 9f9a822509e5ad3e560cbbe830d1013f936fca28 | `incompatible_msrv`: lint function calls with any argument count
The lint for function calls was previously restricted to functions taking
exactly one argument. This was not documented.
Generalizing the lint to an arbitrary number of arguments in the function
call requires special casing some macro expansions from th... | [
{
"path": "clippy_lints/src/incompatible_msrv.rs",
"patch": "@@ -4,12 +4,12 @@ use clippy_utils::is_in_test;\n use clippy_utils::msrvs::Msrv;\n use rustc_attr_parsing::{RustcVersion, StabilityLevel, StableSince};\n use rustc_data_structures::fx::FxHashMap;\n-use rustc_hir::{Expr, ExprKind, HirId};\n+use rus... | 2025-02-13T12:46:30 |
golang/go | 0b71ce6768acd33f188e3edfdc40046c7f6ddf57 | 789d77a87e5417c10377a9f9de07ec37c65048f2 | cmd/go/internal/modload: don't report path errors when loading retractions
When we load module retractions from the latest version of a module,
it's possible that the latest version has a different module path than
the version we're loading. This can happen when a module is renamed or
a go.mod file is added for the fi... | [
{
"path": "src/cmd/go/internal/modload/modfile.go",
"patch": "@@ -122,12 +122,21 @@ func checkRetractions(ctx context.Context, m module.Version) error {\n \n \t\t// Load go.mod for that version.\n \t\t// If the version is replaced, we'll load retractions from the replacement.\n+\t\t//\n \t\t// If there's an... | 2020-09-18T15:48:18 |
ollama/ollama | aabd71aede99443d5858213eec4e8d4c10713c55 | f321b13a032ee287d4ea1139c14c25eed55328d3 | fix rendering and variable width issues on progress bar | [
{
"path": "progress/bar.go",
"patch": "@@ -42,6 +42,19 @@ func NewBar(message string, maxValue, initialValue int64) *Bar {\n \t}\n }\n \n+// formatDuration limits the rendering of a time.Duration to 2 units\n+func formatDuration(d time.Duration) string {\n+\tif d >= 100*time.Hour {\n+\t\treturn \"99h+\"\n+\... | 2023-11-21T15:02:28 |
ggml-org/llama.cpp | 5bd341c9a135a13f901c4cacacc27fa5b299ce19 | 1c7cf94b22a9dc6b1d32422f72a627787a4783a3 | CUDA: Fix builds for older CCCL versions by ifdefing strided_iterator (#18964)
* CUDA: Fix builds for older CCCL versions by ifdefing strided_iterator
Strided iterator was added in [CCCL
3.1](https://github.com/NVIDIA/cccl/releases/tag/v3.1.0), which is packaged into
[CTK
13.1](https://docs.nvidia.com/cuda/cuda-toolk... | [
{
"path": "ggml/src/ggml-cuda/argsort.cu",
"patch": "@@ -2,6 +2,9 @@\n \n #ifdef GGML_CUDA_USE_CUB\n # include <cub/cub.cuh>\n+# if (CCCL_MAJOR_VERSION >= 3 && CCCL_MINOR_VERSION >= 1)\n+# define STRIDED_ITERATOR_AVAILABLE\n+# endif\n using namespace cub;\n #endif // GGML_CUDA_USE_CUB\n \n@... | 2026-01-21T01:34:29 |
nodejs/node | a627c5fc136b3233959e8e1bff90ae4b2d0ec322 | aaf2a1c2264fffd6eea82fe3778f41ac046e6349 | net: fix timeouts during long writes
Add updateWriteQueueSize which updates and returns queue size
(net & tls). Make _onTimeout check whether an active write
is ongoing and if so, call _unrefTimer rather than emitting
a timeout event.
Add http & https test that checks whether long-lasting (but
active) writes timeout ... | [
{
"path": "lib/net.js",
"patch": "@@ -397,6 +397,14 @@ Socket.prototype.setTimeout = function(msecs, callback) {\n \n \n Socket.prototype._onTimeout = function() {\n+ // `.prevWriteQueueSize` !== `.updateWriteQueueSize()` means there is\n+ // an active write in progress, so we suppress the timeout.\n+ co... | 2017-10-12T13:57:42 |
huggingface/transformers | 12c5544dca0fbef4d85c9dbe7aefef29b86b7905 | fbabd6746f7f129adc66f0290bb45ecbfe94236c | Fix memory leak with CTC training script on Chinese languages (#30358)
* Fix memory leak with CTC training script on Chinese languages
* Fix lint | [
{
"path": "examples/pytorch/speech-recognition/run_speech_recognition_ctc.py",
"patch": "@@ -28,7 +28,6 @@\n \n import datasets\n import evaluate\n-import numpy as np\n import torch\n from datasets import DatasetDict, load_dataset\n \n@@ -712,10 +711,14 @@ def is_audio_in_length_range(length):\n log... | 2024-05-02T08:33:36 |
ollama/ollama | 8c4022b06b4ff593439ed8ef29e8362f3844172c | 433702f4218c81c1ef2565b5f765172cc2f90e96 | fix initial progress stats | [
{
"path": "progress/bar.go",
"patch": "@@ -124,15 +124,14 @@ func (b *Bar) Stats() Stats {\n \n \tswitch {\n \tcase b.statted.IsZero():\n-\tcase b.currentValue >= b.maxValue:\n \t\tb.stats = Stats{\n-\t\t\tvalue: b.maxValue,\n+\t\t\tvalue: b.initialValue,\n \t\t\trate: 0,\n \t\t\tremaining: 0,\... | 2023-11-20T19:33:46 |
golang/go | bf9800c7931da96e2ddc8363955f43581265b95c | c47ba5fbfb86edbf49450b8b7b173ab2d188f41f | cmd/go: fix review comments in load, modload
Follow-up to CL 255719, fixing review comments.
Change-Id: I26d3cc622496c3902c6924cf2b746c50705c4d50
Reviewed-on: https://go-review.googlesource.com/c/go/+/255937
Trust: Jay Conrod <jayconrod@google.com>
Run-TryBot: Jay Conrod <jayconrod@google.com>
TryBot-Result: Go Bot <... | [
{
"path": "src/cmd/go/internal/load/pkg.go",
"patch": "@@ -420,13 +420,16 @@ type ImportPathError interface {\n \tImportPath() string\n }\n \n+var (\n+\t_ ImportPathError = (*importError)(nil)\n+\t_ ImportPathError = (*modload.ImportMissingError)(nil)\n+)\n+\n type importError struct {\n \timportPath string... | 2020-09-18T12:33:44 |
huggingface/transformers | fbabd6746f7f129adc66f0290bb45ecbfe94236c | 5cf3e6bf05264726ffa61709787312b09cb4973e | Fix for Neuron (#30259) | [
{
"path": "src/transformers/models/cohere/modeling_cohere.py",
"patch": "@@ -1010,8 +1010,11 @@ def _update_causal_mask(\n causal_mask = causal_mask.clone() # copy to contiguous memory for in-place edit\n if attention_mask.dim() == 2:\n mask_length = attention_mask.s... | 2024-05-02T08:24:47 |
nodejs/node | aaf2a1c2264fffd6eea82fe3778f41ac046e6349 | fe13e0077f5923e327465c9cd00d344cefee800f | tls: properly track writeQueueSize during writes
Make writeQueueSize represent the actual size of the write queue
within the TLS socket. Add tls test to confirm that bufferSize
works as expected.
PR-URL: https://github.com/nodejs/node/pull/15791
Fixes: https://github.com/nodejs/node/issues/15005
Refs: https://github.... | [
{
"path": "lib/_tls_wrap.js",
"patch": "@@ -455,9 +455,8 @@ TLSSocket.prototype._init = function(socket, wrap) {\n var ssl = this._handle;\n \n // lib/net.js expect this value to be non-zero if write hasn't been flushed\n- // immediately\n- // TODO(indutny): revise this solution, it might be 1 before ... | 2017-10-05T20:16:20 |
ggml-org/llama.cpp | 2c1f1996535cd49132e6e7d1b28908bcd4f56819 | d1e3556481c8b351f9b7b69ba3febf6cb77fffa6 | cli : fix reasoning responses in CLI (#18961)
* cli : fix reasoning responses in CLI
* fix build
* fix build (2) | [
{
"path": "common/chat-parser.cpp",
"patch": "@@ -129,7 +129,7 @@ static void parse_json_tool_calls(\n }\n }\n \n-common_chat_msg_parser::common_chat_msg_parser(const std::string & input, bool is_partial, const common_chat_syntax & syntax)\n+common_chat_msg_parser::common_chat_msg_parser(const std::stri... | 2026-01-20T17:23:25 |
ollama/ollama | be61a81758ecd7990e0bd4cdd77093ffc4ecb1ca | 2fdf1b5ff891383752593813e5133f815819f271 | main-gpu argument is not getting passed to llamacpp, fixed. (#1192) | [
{
"path": "llm/llama.go",
"patch": "@@ -339,6 +339,7 @@ func newLlama(model string, adapters []string, runners []ModelRunner, numLayers\n \t\t\"--model\", model,\n \t\t\"--ctx-size\", fmt.Sprintf(\"%d\", opts.NumCtx),\n \t\t\"--batch-size\", fmt.Sprintf(\"%d\", opts.NumBatch),\n+\t\t\"--main-gpu\", fmt.Spri... | 2023-11-20T15:52:52 |
golang/go | df73945fd2fa0b7c168a042e87e648fdfdfc2c70 | 4ffc2bc533ed39b3cbb343ad5873105bfd58ff10 | cmd/compile: make error message involving variadic calls clearer
Fixes #41440
Change-Id: I2fbac72ae3b76bca32cdeaee678a19af3595d116
Reviewed-on: https://go-review.googlesource.com/c/go/+/255241
Trust: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Go Bot ... | [
{
"path": "src/cmd/compile/internal/gc/typecheck.go",
"patch": "@@ -2717,7 +2717,7 @@ func errorDetails(nl Nodes, tstruct *types.Type, isddd bool) string {\n // sigrepr is a type's representation to the outside world,\n // in string representations of return signatures\n // e.g in error messages about wrong... | 2020-09-17T00:54:01 |
huggingface/transformers | 5cf3e6bf05264726ffa61709787312b09cb4973e | c681b58b06f6fb8b5c331f380548af3b4b33f881 | Fix: failing CI after #30568 (#30599)
* failiing CI
* no let's keep it intil full deprecation in v4.42 | [
{
"path": "src/transformers/modeling_utils.py",
"patch": "@@ -1905,6 +1905,9 @@ def resize_token_embeddings(\n return model_embeds\n \n # Update base model and current model config\n+ if hasattr(self.config, \"text_config\"):\n+ self.config.text_config.vocab_size = mode... | 2024-05-02T07:15:17 |
ggml-org/llama.cpp | 959ecf7f234dc0bc0cd6829b25cb0ee1481aa78a | 4037093c66bfe53b9b27d72765416815fdb0398f | jinja : fix undefined keys and attributes and int/float as bool (#18924)
* fix undefined keys and attributes
* add falsy tests
* as_bool for integers and floats
* more falsy/truthy tests
* --typo | [
{
"path": "common/jinja/runtime.cpp",
"patch": "@@ -805,7 +805,7 @@ value member_expression::execute_impl(context & ctx) {\n } else if (is_val<value_string>(property)) {\n auto key = property->as_string().str();\n JJ_DEBUG(\"Accessing %s built-in '%s'\", is_val<value_array>(o... | 2026-01-19T19:29:43 |
nodejs/node | 5018156628a596fff04681826bf2d45423ef4ba1 | 353040c824b038bf4e4c37a669be7e448a7f387f | build,win: use /MP for debug builds
PR-URL: https://github.com/nodejs/node/pull/16333
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Tobias Nießen <tniessen@tnie.de>
Reviewed-By: Refael Ackermann <refack@gmail.com... | [
{
"path": "common.gypi",
"patch": "@@ -175,9 +175,6 @@\n 'EnableFunctionLevelLinking': 'true',\n 'EnableIntrinsicFunctions': 'true',\n 'RuntimeTypeInfo': 'false',\n- 'AdditionalOptions': [\n- '/MP', # compile across multiple CPUs\n- ],\n... | 2017-10-20T08:15:00 |
ollama/ollama | 6bbd6e26fb81637de90779760bf744f4dc3ad8b2 | e6ad4813d3ae1ac229aed71c5e609f829f61a25d | fix temporary newline created and removed with spinner in `ollama run` | [
{
"path": "cmd/cmd.go",
"patch": "@@ -457,7 +457,7 @@ func generate(cmd *cobra.Command, model, prompt string, wordWrap bool, format st\n \t}\n \n \tp := progress.NewProgress(os.Stderr)\n-\tdefer p.Stop()\n+\tdefer p.StopAndClear()\n \n \tspinner := progress.NewSpinner(\"\")\n \tp.Add(\"\", spinner)\n@@ -492... | 2023-11-20T05:49:08 |
rust-lang/rust | 4c1f51bf6ecc68fb6eb4b90ea6b0d0b027146aee | 60493b8973ac5ba632952eaa2f212b56bb97ccfe | Fix link failure on AVR (incompatible ISA error)
Fixes #137739. A reproducer of the issue is present there. I believe the
root cause was introducing the avr-none target (which has no CPU by
default) and trying to get the ISA revision from there. This commit
uses the `target-cpu` option instead, which is already requir... | [
{
"path": "compiler/rustc_codegen_ssa/src/back/metadata.rs",
"patch": "@@ -373,7 +373,11 @@ pub(crate) fn create_object_file(sess: &Session) -> Option<write::Object<'static\n Architecture::Avr => {\n // Resolve the ISA revision and set\n // the appropriate EF_AVR_ARCH flag.\n... | 2025-02-28T17:24:16 |
nodejs/node | f8063d51d7e83ab6d5c7cfcadc37ed6b4f6d8ef5 | 2146c88bc7383ce66839e08cc22da17a51f5cacf | benchmark: fix punycode test for --without-intl
PR-URL: https://github.com/nodejs/node/pull/16251
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com> | [
{
"path": "benchmark/misc/punycode.js",
"patch": "@@ -1,11 +1,14 @@\n 'use strict';\n \n const common = require('../common.js');\n-const icu = process.binding('icu');\n+let icu;\n+try {\n+ icu = process.binding('icu');\n+} catch (err) {}\n const punycode = require('punycode');\n \n const bench = common.cre... | 2017-10-17T05:31:52 |
rust-lang/rust | adb5ecabdb09ff6f329bbf9b7721036db983f546 | f45d4acf1bb635aa010f19f8a749eed8293203b3 | Tweak invalid RTN errors
Make suggestions verbose.
When encountering `method(type)` bound, suggest `method(..)` instead of `method()`.
```
error: argument types not allowed with return type notation
--> $DIR/bad-inputs-and-output.rs:9:23
|
LL | fn foo<T: Trait<method(i32): Send>>() {}
| ... | [
{
"path": "compiler/rustc_ast_lowering/messages.ftl",
"patch": "@@ -37,11 +37,11 @@ ast_lowering_bad_return_type_notation_inputs =\n .suggestion = remove the input types\n \n ast_lowering_bad_return_type_notation_needs_dots = return type notation arguments must be elided with `..`\n- .suggestion = ad... | 2025-02-28T20:05:43 |
ggml-org/llama.cpp | 18361c579cc7ed0a06ed9085eaf900326b537fa7 | 365a3e8c319ddb5442afdf17d0d23dfa0ff26c78 | server: fix memory reservations in populate_token_probs (#18787) | [
{
"path": "tools/server/server-context.cpp",
"patch": "@@ -1326,11 +1326,12 @@ struct server_context_impl {\n }\n \n void populate_token_probs(const server_slot & slot, completion_token_output & result, bool post_sampling, bool special, int idx) const {\n- const size_t n_probs = slot.task->pa... | 2026-01-19T18:13:31 |
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.