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
electron/electron
59246a4c7c27253848fd6c56bc7c9a137b85f123
2205d725f2b01e5cc78e7e38fb597d53285c15ac
fix: NOTREACHED when resizing windows frameless bounds (#33029) * fix: NOTREACHED when resizing windows frameless bounds * fix: scope to window_->IsResizable()
[ { "path": "shell/browser/ui/views/win_frame_view.cc", "patch": "@@ -34,6 +34,14 @@ void WinFrameView::Init(NativeWindowViews* window, views::Widget* frame) {\n window_ = window;\n frame_ = frame;\n \n+ // Prevent events from trickling down the views hierarchy here, since\n+ // when a given resizable w...
2022-03-21T23:30:05
ggml-org/llama.cpp
cad2d3884c75e3b2e4736654295d1f1a5c29aa48
389c7d4955ba55c7418afaebf7c23d9ed64ef707
rpc : fix misleading error log (#21184) When RPC is running with a remote backend which doesn't have init_tensor function (like CPU and Metal), the server log gets full with error messages saying that init_tensor is being called with null buffer which is incorrect. This patch fixes this.
[ { "path": "ggml/src/ggml-rpc/ggml-rpc.cpp", "patch": "@@ -1340,7 +1340,9 @@ bool rpc_server::init_tensor(const rpc_msg_init_tensor_req & request) {\n if (buffer && buffer->iface.init_tensor) {\n buffer->iface.init_tensor(buffer, tensor);\n } else {\n- GGML_LOG_ERROR(\"Null buffer for ...
2026-03-30T14:05:11
facebook/react
18282f881dae106ebf6240aa52c8c02fe7c8d6f2
540bab085d571789f4562565eebfd0db9f36345c
Fix: Update while suspended fails to interrupt (#26739) This fixes a bug with `use` where if you update a component that's currently suspended, React will sometimes mistake it for a render phase update. This happens because we don't reset `currentlyRenderingFiber` until the suspended is unwound. And with `use`, ...
[ { "path": "packages/react-reconciler/src/ReactFiberHooks.js", "patch": "@@ -711,6 +711,9 @@ function renderWithHooksAgain<Props, SecondArg>(\n //\n // Keep rendering in a loop for as long as render phase updates continue to\n // be scheduled. Use a counter to prevent infinite loops.\n+\n+ currentlyRe...
2023-04-28T19:35:14
golang/go
7b296625e478658cfbd43fd09e30dde01992ba6f
0751f3af923af6ab5330b88a8c00ac2ddeddedb6
doc/next: add notes for new #cgo annotations Fixes #56378 Change-Id: Ibeaa5aa67e017be0da8797e4c87abf2ee0249db8 Reviewed-on: https://go-review.googlesource.com/c/go/+/632015 LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com> Reviewed-by: Ian Lance Taylor <iant@google.com> Reviewe...
[ { "path": "doc/next/3-tools.md", "patch": "@@ -14,6 +14,14 @@ a test integration system, you can revert to the text build output by setting\n \n ### Cgo {#cgo}\n \n+Cgo supports new annotations for C functions to improve run time\n+performance.\n+`#cgo noescape cFunctionName` tells the compiler that memory ...
2024-11-26T20:52:31
huggingface/transformers
e6fa62bf8e60312d1275ee34a06efd80f0389aa8
6c6f70cb9da270c871d1a2177cf7af90faa5c418
fix: TextClassificationPipeline docs mentioning deprecated return_all_scores (#43903)
[ { "path": "src/transformers/pipelines/text_classification.py", "patch": "@@ -30,8 +30,6 @@ class ClassificationFunction(ExplicitEnum):\n @add_end_docstrings(\n build_pipeline_init_args(has_tokenizer=True),\n r\"\"\"\n- return_all_scores (`bool`, *optional*, defaults to `False`):\n- ...
2026-02-11T16:59:23
ollama/ollama
6bf0b8193acadc5d595f320796d9876302ae0620
db428adbb81413d8be3933eb0ec2256ab36880ce
readme: fix typos (#10399)
[ { "path": "README.md", "patch": "@@ -285,7 +285,7 @@ See the [API documentation](./docs/api.md) for all endpoints.\n - [Bionic GPT](https://github.com/bionic-gpt/bionic-gpt)\n - [HTML UI](https://github.com/rtcfirefly/ollama-ui)\n - [Saddle](https://github.com/jikkuatwork/saddle)\n-- [TagSpaces](https://www...
2025-04-29T17:30:44
electron/electron
4cc2ed842e6ee6bab9c9b8743593023953da046b
e100402b1335315d707c72b69cd940ad5aee5c15
fix: iocp integration when process is reused (#33207)
[ { "path": "shell/common/node_bindings.cc", "patch": "@@ -462,7 +462,8 @@ node::Environment* NodeBindings::CreateEnvironment(\n \n args.insert(args.begin() + 1, init_script);\n \n- isolate_data_ = node::CreateIsolateData(isolate, uv_loop_, platform);\n+ if (!isolate_data_)\n+ isolate_data_ = node::Cre...
2022-03-21T07:42:22
facebook/react
d2f33dd524802245ce111917e4fd764b2657042f
d5dee809338c551a8ff2301eabdfc9e83609eb91
Retain locations when using babel cloneNode in lambdas It turns out the third parameter to `cloneNode` is ["If the third parameter is true, the cloned nodes exclude location properties."](https://github.com/babel/babel/blob/c060e5e3d538a174a017edb7e75b07e80ccb55b7/packages/babel-types/src/clone/cloneNode.ts#L35-L39...
[ { "path": "compiler/forget/src/ReactiveScopes/CodegenReactiveFunction.ts", "patch": "@@ -895,7 +895,7 @@ function codegenInstructionValue(\n break;\n }\n case \"FunctionExpression\": {\n- value = t.cloneNode(instrValue.expr, true, true);\n+ value = t.cloneNode(instrValue.expr, true, ...
2023-04-27T20:25:41
ggml-org/llama.cpp
389c7d4955ba55c7418afaebf7c23d9ed64ef707
278521c33a11b89d9d7ed2afe5c20502840816b1
webui: Fix branching logic on edit message (#21175) * fix: Branching logic + small refactor * chore: update webui build output
[ { "path": "tools/server/webui/src/lib/stores/chat.svelte.ts", "patch": "@@ -28,6 +28,7 @@ import {\n \tfilterByLeafNodeId,\n \tfindDescendantMessages,\n \tfindLeafNode,\n+\tfindMessageById,\n \tisAbortError\n } from '$lib/utils';\n import {\n@@ -416,7 +417,7 @@ class ChatStore {\n \t\tif (!activeConv) retur...
2026-03-30T12:40:50
golang/go
0751f3af923af6ab5330b88a8c00ac2ddeddedb6
cff2503b5ac32d3889eec8f131c43e98e4cc3485
doc/next: add release note for vet on Printf(s) with non-const s Fixes #60529 Updates #68545 Change-Id: Ica048e4861f4de22deb7f42a1d76eb3151fee986 Reviewed-on: https://go-review.googlesource.com/c/go/+/631682 Auto-Submit: Alan Donovan <adonovan@google.com> Reviewed-by: Robert Findley <rfindley@google.com> LUCI-TryBot-...
[ { "path": "doc/next/3-tools.md", "patch": "@@ -28,9 +28,14 @@ tests, fuzzers, benchmarks, and examples in test packages, such as\n malformed names, incorrect signatures, or examples that document\n non-existent identifiers. Some of these mistakes may cause tests not\n to run.\n-\n This analyzer is among the...
2024-11-26T20:31:36
ollama/ollama
6ed88985903be474ecd59992f7191c2f0fa87e36
5cfc1c39f3d5822b0c0906f863f6df45c141c33b
ggml: fix crash for array head counts If it's an array, it uses the max value in the array If array values for head counts becomes more popular, we can consider a more invasive change like #10225 to calculate more accurate estimates. Fixes: #9984
[ { "path": "fs/ggml/ggml.go", "patch": "@@ -33,7 +33,8 @@ func (kv KV) Kind() string {\n }\n \n func (kv KV) ParameterCount() uint64 {\n-\treturn keyValue(kv, \"general.parameter_count\", uint64(0))\n+\tval, _ := keyValue(kv, \"general.parameter_count\", uint64(0))\n+\treturn val\n }\n \n func (kv KV) FileTy...
2025-04-25T23:16:15
facebook/react
d5dee809338c551a8ff2301eabdfc9e83609eb91
3eed52b49b0885840ca236fee18724f3564b47e8
Failing test for fbt issue Specifically, the fbt plugin would error on a `null` loc as it [relies on it in one of its utils](https://github.com/facebook/fbt/blob/6a23a5374a757a49654d8304a3efea8eb9548eea/packages/babel-plugin-fbt/src/getNamespacedArgs.js#L82-L84), implying a missing `loc` somewhere in our pipeline ...
[ { "path": "compiler/forget/src/__tests__/fixtures/compiler/error.lambda-with-fbt.expect.md", "patch": "@@ -0,0 +1,45 @@\n+\n+## Input\n+\n+```javascript\n+import { fbt } from \"fbt\";\n+\n+function Component() {\n+ const buttonLabel = () => {\n+ if (!someCondition) {\n+ return <fbt desc=\"My label\...
2023-04-27T20:25:41
ggml-org/llama.cpp
e2eb39e81cbd43ff95dfc46bddb5155b582ccdfd
abf9a62161f232cfb88509005d2e6b65774d9a3e
ci : bump ty to 0.0.26 (#21156) * fix incorrect type ignore comments * bump ty to 0.0.26
[ { "path": ".github/workflows/python-type-check.yml", "patch": "@@ -31,7 +31,7 @@ jobs:\n uses: actions/setup-python@v6\n with:\n python-version: \"3.11\"\n- pip-install: -r requirements/requirements-all.txt ty==0.0.24\n+ pip-install: -r requirements/requirements-a...
2026-03-30T07:29:15
electron/electron
45e2f86fe018f3df8aad56e1a674ff2b50cd1c9b
fdb60240f31b70130cbbf2d40e679f09c5919737
chore: fix pylint-2.7 errors (#33233)
[ { "path": "script/add-debug-link.py", "patch": "@@ -14,8 +14,7 @@ def add_debug_link_into_binaries(directory, target_cpu, debug_dir):\n add_debug_link_into_binary(binary_path, target_cpu, debug_dir)\n \n def add_debug_link_into_binary(binary_path, target_cpu, debug_dir):\n- if PLATFORM == 'linux' and...
2022-03-21T02:11:21
golang/go
cff2503b5ac32d3889eec8f131c43e98e4cc3485
04879acdebbb08bdca00356f043d769c4b4375ce
doc: add note about x509sha1 removal Fixes #41682 Change-Id: I73aa4991a6df8dfb90720bd66f2a8a53c3ca3504 Reviewed-on: https://go-review.googlesource.com/c/go/+/631677 Reviewed-by: Dmitri Shuralyov <dmitshur@google.com> Reviewed-by: Dmitri Shuralyov <dmitshur@golang.org> Auto-Submit: Roland Shoemaker <roland@golang.org>...
[ { "path": "doc/next/6-stdlib/99-minor/crypto/x509/41682.md", "patch": "@@ -0,0 +1,2 @@\n+The `x509sha1` GODEBUG setting has been removed. [`Certificate.Verify`] will no\n+longer consider SHA-1 based signatures valid when this GODEBUG setting is set.\n\\ No newline at end of file", "additions": 2, "d...
2024-11-26T18:18:50
ollama/ollama
5cfc1c39f3d5822b0c0906f863f6df45c141c33b
f0ad49ea17d587cce7f4b2c6a6ccb3139ec083c8
model: fix build (#10416)
[ { "path": "model/models/llama4/model.go", "patch": "@@ -39,7 +39,7 @@ func New(c fs.Config) (model.Model, error) {\n \t\t\t\t`[^\\r\\n\\p{L}\\p{N}]?[\\p{Lu}\\p{Lt}\\p{Lm}\\p{Lo}\\p{M}]*[\\p{Ll}\\p{Lm}\\p{Lo}\\p{M}]+(?i:'s|'t|'re|'ve|'m|'ll|'d)?|[^\\r\\n\\p{L}\\p{N}]?[\\p{Lu}\\p{Lt}\\p{Lm}\\p{Lo}\\p{M}]+[\\p...
2025-04-26T02:24:48
huggingface/transformers
6c6f70cb9da270c871d1a2177cf7af90faa5c418
ac6cba6687a877bef09114d23117ea18f9f6b298
Harmonize `input_embeds` to `inputs_embeds` everywhere (#43916) * fix all * fix repo * oupsi * need to switch decorator order
[ { "path": "src/transformers/generation/utils.py", "patch": "@@ -502,10 +502,10 @@ def _cache_dependant_input_preparation(\n as it needs a different implementation to be exportable.\n \n If we have cache: let's slice `input_ids` through `cache_position`, to keep only the unprocessed tokens\n-...
2026-02-11T16:03:51
electron/electron
fdb60240f31b70130cbbf2d40e679f09c5919737
08d54d241693441ef672efa87009c0eb01ae86cb
fix: removed unused imported Objects (#33304)
[ { "path": "script/lib/util.py", "patch": "@@ -1,18 +1,13 @@\n #!/usr/bin/env python\n \n from __future__ import print_function\n-import atexit\n import contextlib\n import errno\n import json\n import os\n import shutil\n-import ssl\n-import stat\n import subprocess\n import sys\n-import tarfile\n-import te...
2022-03-21T01:21:21
ggml-org/llama.cpp
abf9a62161f232cfb88509005d2e6b65774d9a3e
7c203670f8d746382247ed369fea7fbf10df8ae0
server: wrap headers for mcp proxy (#21072) * server: wrap headers for mcp proxy * Update tools/server/server-cors-proxy.h Co-authored-by: Georgi Gerganov <ggerganov@gmail.com> * fix build * chore: update webui build output * chore: update webui build output --------- Co-authored-by: Georgi Gerganov <ggerganov@...
[ { "path": "tools/server/server-cors-proxy.h", "patch": "@@ -32,13 +32,22 @@ static server_http_res_ptr proxy_request(const server_http_req & req, std::strin\n \n SRV_INF(\"proxying %s request to %s://%s:%i%s\\n\", method.c_str(), parsed_url.scheme.c_str(), parsed_url.host.c_str(), parsed_url.port, parse...
2026-03-30T06:59:16
ollama/ollama
7ba9fa9c7d0bc73abacca88d6827d973d7ba92cf
8bf11b84c1870b88e9e1568041ee837a062aa2c0
fixes for maverick
[ { "path": "convert/convert_llama4.go", "patch": "@@ -45,8 +45,8 @@ func (p *llama4Model) KV(t *Tokenizer) ggml.KV {\n \t\t}\n \t}\n \n-\tkv[\"llama4.intermediate_size\"] = p.TextModel.IntermediateSizeMLP\n-\tkv[\"llama4.intermediate_size_moe\"] = p.TextModel.IntermediateSize\n+\tkv[\"llama4.feed_forward_len...
2025-04-21T17:45:56
golang/go
04879acdebbb08bdca00356f043d769c4b4375ce
592da0ba474b94b6eceee62b5613f1c9c1ed9c89
net/http: test for racing idle conn closure and new requests TestTransportRemovesH2ConnsAfterIdle is experiencing flaky failures due to a bug in idle connection handling. Upon inspection, TestTransportRemovesH2ConnsAfterIdle is slow and (I think) not currently testing the condition that it was added to test. Using th...
[ { "path": "src/net/http/clientserver_test.go", "patch": "@@ -40,9 +40,10 @@ import (\n type testMode string\n \n const (\n-\thttp1Mode = testMode(\"h1\") // HTTP/1.1\n-\thttps1Mode = testMode(\"https1\") // HTTPS/1.1\n-\thttp2Mode = testMode(\"h2\") // HTTP/2\n+\thttp1Mode = testMode(\"...
2024-11-25T19:27:33
huggingface/transformers
ac6cba6687a877bef09114d23117ea18f9f6b298
48295ecbf7d005584815e69e103cfb72aa4e079a
Ernie4 5 vl moe (#43755) * fix unsteady output for ernie4_5_vl_moe Signed-off-by: Liu, Kaixuan <kaixuan.liu@intel.com> * update Signed-off-by: Liu, Kaixuan <kaixuan.liu@intel.com> * add pad_side=left Signed-off-by: Liu, Kaixuan <kaixuan.liu@intel.com> * fixup --> changed on the hub * update Expectations Signed...
[ { "path": "tests/models/ernie4_5_vl_moe/test_modeling_ernie4_5_vl_moe.py", "patch": "@@ -25,7 +25,9 @@\n is_vision_available,\n )\n from transformers.testing_utils import (\n+ Expectations,\n cleanup,\n+ require_deterministic_for_xpu,\n require_torch,\n require_torch_large_accelerator,...
2026-02-11T15:53:28
ggml-org/llama.cpp
f5d1c4179fedf726bec744d3125a55df8d02496a
2405d59cb613f7b9f98ecbc9eb25f8a45188ee06
hexagon: dma optimizations (mostly fixing regressions) (#21137) * hex-fa: add simple dma cache for Mask I noticed that we were refetch the mask rows over and over. This simple cache avoids that. * hex-dma: unset in-order desc bit which caused signficant perf regression We don't rely on true in order processing of t...
[ { "path": "ggml/src/ggml-hexagon/htp/flash-attn-ops.c", "patch": "@@ -346,6 +346,9 @@ static void flash_attn_ext_f16_thread(unsigned int nth, unsigned int ith, void *\n \n const HVX_Vector logit_cap = hvx_vec_splat_f32(factx->logit_softcap);\n \n+ dma_cache m_cache;\n+ dma_cache_init(&m_cache, spa...
2026-03-29T13:40:13
electron/electron
08d54d241693441ef672efa87009c0eb01ae86cb
e07c2b84d76e649478e44a81a445c4076c4e603f
fix: intialize FPS file in network service (#33339)
[ { "path": "shell/browser/net/system_network_context_manager.cc", "patch": "@@ -288,6 +288,10 @@ void SystemNetworkContextManager::OnNetworkServiceCreated(\n base::FeatureList::IsEnabled(features::kAsyncDns),\n default_secure_dns_mode, doh_config, additional_dns_query_types_enabled);\n \n+ // In...
2022-03-19T04:29:42
ollama/ollama
54055a6dae0588d178eb355a8c36051d1b6e98a5
340448d2d1689f707d7644bc8ca018e0db015c9c
fix test
[ { "path": "sample/samplers_test.go", "patch": "@@ -74,7 +74,6 @@ func modelHelper(t testing.TB) model.BytePairEncoding {\n \t\tt.Fatal(err)\n \t}\n \n-\ttypes := make([]uint32, len(vocab))\n \ttokens := make([]string, len(vocab))\n \tfor token, id := range vocab {\n \t\ttokens[id] = token\n@@ -86,7 +85,7 @@...
2025-04-25T23:15:08
golang/go
b68f8ca89a2fb6b33d1e78fadc33d1c35693f6b8
8397de2241ca71ba5df2fa60564a8559b9a2d25d
crypto/internal: keep fips140/aes.NewCTR from allocating Return a *CTR from an always-inlineable function, so the allocation can be lifted to the callsite. Put the potentially uninlineable code in a separate function that returns a CTR. Fixes #70499 Change-Id: I2531a2516e4c00aba65407f3bc24a7c21dd8f842 Reviewed-on: ...
[ { "path": "src/crypto/internal/fips140/aes/ctr.go", "patch": "@@ -18,11 +18,18 @@ type CTR struct {\n }\n \n func NewCTR(b *Block, iv []byte) *CTR {\n+\t// Allocate the CTR here, in an easily inlineable function, so\n+\t// the allocation can be done in the caller's stack frame\n+\t// instead of the heap. S...
2024-11-22T23:19:35
huggingface/transformers
64e41924f45d37593c8297b50578f432b6f893da
b52b6631a26fd362e49968af14ed77ff359006a0
Move `_keys_to_ignore_on_load_missing` for now (#43893) * should be under pretrained model for now * fix * modular fixes * more * last? * update modulars * isolate? * last nit
[ { "path": "src/transformers/models/deepseek_v3/modeling_deepseek_v3.py", "patch": "@@ -550,6 +550,7 @@ class DeepseekV3PreTrainedModel(PreTrainedModel):\n \"attentions\": DeepseekV3Attention,\n }\n _keep_in_fp32_modules_strict = [\"e_score_correction_bias\"]\n+ _keys_to_ignore_on_load_une...
2026-02-11T13:47:39
ggml-org/llama.cpp
65097181e4c8358e7b570b48a2ae7b759f12bf26
98ae0a0d3600f08fbd8d938bc8de0436755e50c3
fix **/x glob matching (#21129)
[ { "path": "common/common.cpp", "patch": "@@ -703,7 +703,6 @@ static inline bool glob_match(const char * pattern, const char * str) {\n }\n if (pattern[0] == '*' && pattern[1] == '*') {\n const char * p = pattern + 2;\n- if (*p == '/') p++;\n if (glob_match(p, str)) return true...
2026-03-28T21:27:38
ollama/ollama
ced7d0e53df146200c322ccc7f3493aa32f627e1
a0dba0f8aefad9a843b59040857f4a04021c54e1
fix parameter count
[ { "path": "fs/ggml/ggml.go", "patch": "@@ -33,7 +33,7 @@ func (kv KV) Kind() string {\n }\n \n func (kv KV) ParameterCount() uint64 {\n-\treturn keyValue[uint64](kv, \"general.parameter_count\")\n+\treturn keyValue(kv, \"general.parameter_count\", uint64(0))\n }\n \n func (kv KV) FileType() fileType {", ...
2025-04-23T23:07:11
electron/electron
e07c2b84d76e649478e44a81a445c4076c4e603f
f2b06324b8836a4a3acc93c86f77a624c04b855b
fix: ensure ElectronBrowser mojo service is only bound to appropriate render frames (#33323) * fix: ensure ElectronBrowser mojo service is only bound to authorized render frames Notes: no-notes * refactor: extract electron API IPC to its own mojo interface * fix: just check main frame not primary main frame
[ { "path": "filenames.gni", "patch": "@@ -350,6 +350,8 @@ filenames = {\n \"shell/browser/child_web_contents_tracker.h\",\n \"shell/browser/cookie_change_notifier.cc\",\n \"shell/browser/cookie_change_notifier.h\",\n+ \"shell/browser/electron_api_ipc_handler_impl.cc\",\n+ \"shell/browser/el...
2022-03-19T02:50:05
golang/go
8fb6a469472b3d79798da0b3f84b9579fae18ec3
d0799a5fe27b779978fc2031fce82904d1242443
cmd/go: pass os.Args[0] to mod tools Fixes #70509 Change-Id: I62952c0ac5de456742b7bee3552a0d90238bfde5 Reviewed-on: https://go-review.googlesource.com/c/go/+/631075 Reviewed-by: Michael Matloob <matloob@golang.org> LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserviceaccount.com> Reviewed-by: ...
[ { "path": "src/cmd/go/internal/tool/tool.go", "patch": "@@ -305,7 +305,7 @@ func runBuiltTool(b *work.Builder, ctx context.Context, a *work.Action) error {\n \n \ttoolCmd := &exec.Cmd{\n \t\tPath: cmdline[0],\n-\t\tArgs: cmdline[1:],\n+\t\tArgs: cmdline,\n \t\tStdin: os.Stdin,\n \t\tStdout: os.Stdout...
2024-11-22T04:03:29
facebook/react
3eed52b49b0885840ca236fee18724f3564b47e8
b69d70664cf52bac084ea092a1cf2d181239aa5e
Only debug with @debug pragma You don't always want to debug all the passes when you use @only, so split it out into its own pragma
[ { "path": "compiler/forget/src/__tests__/test-utils/generateTestsFromFixtures.ts", "patch": "@@ -98,11 +98,13 @@ export default function generateTestsFromFixtures(\n const lines = input.split(\"\\n\");\n if (lines[0]!.indexOf(\"@only\") !== -1) {\n testCommand = test.only;\n-...
2023-04-27T20:16:50
huggingface/transformers
b52b6631a26fd362e49968af14ed77ff359006a0
7d4c298b1ba2ea6c1eda2838321f042c3cb4359e
Fix old tech stack in doc (#43902) Signed-off-by: Yuanyuan Chen <cyyever@outlook.com>
[ { "path": "docs/source/ar/installation.md", "patch": "@@ -2,11 +2,9 @@\n \n قم بتثبيت مكتبة 🤗 Transformers المناسبة لمكتبة التعلم العميق التي تستخدمها، وقم بإعداد ذاكرة التخزين المؤقت الخاصة بك، وقم بإعداد 🤗 Transformers للعمل دون اتصال بالإنترنت (اختياري).\n \n-تم اختبار 🤗 Transformers على Python 3.6 و...
2026-02-11T12:57:10
ggml-org/llama.cpp
98ae0a0d3600f08fbd8d938bc8de0436755e50c3
3a14a542f5ce8666713c6e6ea44f7f3e01dd6e45
common/parser: fix handling of tool definition with missing properties key (#21128)
[ { "path": "common/chat-auto-parser-generator.cpp", "patch": "@@ -65,7 +65,7 @@ common_chat_params peg_generator::generate_parser(const common_chat_template &\n data.grammar = build_grammar([&](const common_grammar_builder & builder) {\n foreach_function(inputs.tools, [&](const json ...
2026-03-28T19:41:32
electron/electron
df8fd1b269ace37e7ad62ffa9ac8895bca463ef6
4342b7ff555a1707ab18f7a71da01ae27ef29cd1
build: fix build by removing usage of removed preloadUrl (#33319)
[ { "path": "lib/browser/api/web-contents.ts", "patch": "@@ -571,7 +571,7 @@ const loggingEnabled = () => {\n // Add JavaScript wrappers for WebContents class.\n WebContents.prototype._init = function () {\n const prefs = this.getLastWebPreferences() || {};\n- if (!prefs.nodeIntegration && (prefs.preload !...
2022-03-17T07:13:48
facebook/react
4a8c90a77cff6e931bbc0395b3f6405ccf900ee9
b1eaf88c61dfaf3fedf4f58e45fad4cf519e8049
[babel] Remove unsafe calls to path.stop() Discovered this in a recent attempt at syncing Forget to Meta, it seems that calling path.stop() is unsafe as it appears to have strange behavior in plugins that come after. This resulted in `import type {...}` not being compiled away in the post-babel output which isn...
[ { "path": "compiler/forget/src/Babel/BabelPlugin.ts", "patch": "@@ -171,7 +171,6 @@ export default function ReactForgetBabelPlugin(\n Directive(path) {\n if (hasUseForgetDirective(path.node)) {\n fileHasUseForgetDirective = true;\n- path.stop();\n ...
2023-04-26T21:16:34
golang/go
043708eb08652cfefe9972bcfbf94464a3566920
7f049eac1b9378ecc4dddd43ebedeae0916c0606
crypto/internal/fips140/aes: handle fallback correctly Don't fallthrough to the hardware version if we used the generic version. This might fix the s390x build on the dashboard. (Originally broken on CL 624738.) Change-Id: Idad1f1973a34fc64550ecf0d012651f62bcd6272 Reviewed-on: https://go-review.googlesource.com/c/g...
[ { "path": "src/crypto/internal/fips140/aes/cbc_s390x.go", "patch": "@@ -15,13 +15,15 @@ func cryptBlocksChain(c code, iv, key, dst, src *byte, length int)\n func cryptBlocksEnc(b *Block, civ *[BlockSize]byte, dst, src []byte) {\n \tif b.fallback != nil {\n \t\tcryptBlocksEncGeneric(b, civ, dst, src)\n+\t\tr...
2024-11-22T18:48:45
huggingface/transformers
7d4c298b1ba2ea6c1eda2838321f042c3cb4359e
007bb8c346d4ebfa7a735a346ed48587fa12a2a1
Fix missing fast image patch counter in Glm46V (#43877) * Fix missing fast image patch counter for Glm46V * Fix review follow-ups for GLM46V fast patch counting test * Apply repo consistency fixes * Fix GLM image patch counter in glm4v source and move regression test * style(glm4v): format image processing regress...
[ { "path": "src/transformers/models/ernie4_5_vl_moe/image_processing_ernie4_5_vl_moe_fast.py", "patch": "@@ -190,14 +190,6 @@ def _preprocess(\n data={\"pixel_values\": pixel_values, \"image_grid_thw\": image_grid_thw}, tensor_type=return_tensors\n )\n \n- @auto_docstring\n- def pre...
2026-02-11T12:23:27
ollama/ollama
d26c18e25c493ca55add9713ed151081c8de7ecf
8d376acc9b3e6796a12e6462c530b1bff3dc2f90
fix token type
[ { "path": "fs/config.go", "patch": "@@ -8,6 +8,6 @@ type Config interface {\n \tBool(string, ...bool) bool\n \n \tStrings(string, ...[]string) []string\n-\tUints(string, ...[]uint32) []uint32\n+\tInts(string, ...[]int32) []int32\n \tFloats(string, ...[]float32) []float32\n }", "additions": 1, "delet...
2025-04-23T19:40:05
ggml-org/llama.cpp
e397d3885c40fac0b91bc2784c2d948a259de8a4
e6f2ec01ff808b60f80ffcb46c748f871576c8f3
common/json-schema: fix: handle non-capturing groups (?:...) in JSON schema pattern converter (#21124) The regex-to-grammar converter in _visit_pattern() crashes with SIGSEGV when a JSON schema "pattern" field contains a non-capturing group (?:...). Root cause: when the parser sees '(' followed by '?', it pushes a wa...
[ { "path": "common/json-schema-to-grammar.cpp", "patch": "@@ -416,15 +416,30 @@ class common_schema_converter {\n i++;\n } else if (c == '(') {\n i++;\n- if (i < length) {\n- if (sub_pattern[i] == '?') {\n+ ...
2026-03-28T16:55:38
electron/electron
e9044860766affd5ded7baa3b44da6d4b9fe89e0
b2c5623a13c91a561704b46480b6c3a7a51f2db7
fix: BrowserWindow.fromWebContents should work in browser-window-created (#33257)
[ { "path": "shell/common/gin_helper/trackable_object.h", "patch": "@@ -117,11 +117,11 @@ class TrackableObject : public TrackableObjectBase, public EventEmitter<T> {\n ~TrackableObject() override { RemoveFromWeakMap(); }\n \n void InitWith(v8::Isolate* isolate, v8::Local<v8::Object> wrapper) override {\n...
2022-03-16T23:23:14
facebook/react
b1eaf88c61dfaf3fedf4f58e45fad4cf519e8049
e5f4b3008fcb598c1bd2b15ec0b52f05cd8a3b11
[babel] Ensure only adding import specifier to non-namespace Missed this in the previous PR Test plan: P706162189 (some babel errors) before this PR, P706255523 has no errors
[ { "path": "compiler/forget/src/Babel/BabelPlugin.ts", "patch": "@@ -233,14 +233,7 @@ export default function ReactForgetBabelPlugin(\n }\n },\n ImportDeclaration(importDeclPath) {\n- // Matches `import { /* ... */ } from 'react';`\n- // but not...
2023-04-26T19:16:21
golang/go
a8eddaf758c15404acbfe61af0739b6d099f2067
00709919d09904b17cfe3bfeb35521cbd3fb04f8
runtime, internal/synctest, syscall/js: keep bubble membership in syscalls Propagate synctest bubble membership through syscall/js.Func functions. Avoids panics from cross-bubble channel operations in js syscalls. Fixes #70512 Change-Id: Idbd9f95da8bc4f055a635dfac041359f848dad1a Reviewed-on: https://go-review.google...
[ { "path": "src/go/build/deps_test.go", "patch": "@@ -131,8 +131,12 @@ var depsRules = `\n \n \tunicode !< path;\n \n+\tRUNTIME\n+\t< internal/synctest\n+\t< testing/synctest;\n+\n \t# SYSCALL is RUNTIME plus the packages necessary for basic system calls.\n-\tRUNTIME, unicode/utf8, unicode/utf16\n+\tRUNTIME,...
2024-11-22T01:32:22
huggingface/transformers
007bb8c346d4ebfa7a735a346ed48587fa12a2a1
372c27e71f80e64571ac1149d1708e641d7d44da
fix(moe): Handle dtype mismatch in torch._grouped_mm with autocast (#43839) * fix(moe): Handle dtype mismatch in grouped_mm with autocast torch._grouped_mm is not autocast-enabled, so when using torch.autocast with MoE models like Phi-tiny-MoE, the input tensor may have a different dtype than the weights, causing Run...
[ { "path": "src/transformers/integrations/moe.py", "patch": "@@ -180,12 +180,15 @@ def _grouped_linear(\n Returns:\n `torch.Tensor`: Output tensor of shape (S, output_dim).\n \"\"\"\n- if is_transposed:\n- # (S, input_dim) @ grouped (num_experts, input_dim, output_dim) -> (S, output...
2026-02-11T10:42:45
ggml-org/llama.cpp
edfb440a2f469ea3abaf397627bdc7d40b7777d8
3d66da1809651543001d37f2b6cb3a172adc2557
server : fix processing of multiple back-to-back mtmd chunks (#21107)
[ { "path": "tools/server/server-context.cpp", "patch": "@@ -2493,7 +2493,7 @@ struct server_context_impl {\n bool has_mtmd = false;\n \n // check if we should process the image\n- if (slot.prompt.n_tokens() < slot.task->n_tokens() && input_tokens[slo...
2026-03-28T14:27:36
electron/electron
b2c5623a13c91a561704b46480b6c3a7a51f2db7
652680e8012983b1d8b9118b48297df4c1d9a939
fix: crash when destroying node env with pending promises (#33280) * fix: crash when destroying node env with pending promises * chore: add spec
[ { "path": "shell/renderer/electron_renderer_client.cc", "patch": "@@ -151,17 +151,22 @@ void ElectronRendererClient::WillReleaseScriptContext(\n if (env == node_bindings_->uv_env())\n node_bindings_->set_uv_env(nullptr);\n \n- // Destroy the node environment. We only do this if node support has been...
2022-03-16T17:54:45
ollama/ollama
214a7678eab94a6acf88fd1682f8fe6733ba555d
4892872c184a8fa81baaedf9669bf94e4a278964
fix superfluous call to WriteHeader the first call to http.ResponseWriter.Write implicitly calls WriteHeader with http.StatusOK if it hasn't already been called. once WriteHeader has been called, subsequent calls has no effect. Write is called when JSON encoding progressUpdateJSON{}. calls to http.ResponseWriter.Write...
[ { "path": "server/internal/registry/server.go", "patch": "@@ -73,8 +73,13 @@ type statusCodeRecorder struct {\n func (r *statusCodeRecorder) WriteHeader(status int) {\n \tif r._status == 0 {\n \t\tr._status = status\n+\t\tr.ResponseWriter.WriteHeader(status)\n \t}\n-\tr.ResponseWriter.WriteHeader(status)\n+...
2025-04-24T20:09:39
facebook/react
2aa3a284810424bee7064fa0748c260f8121bb5e
5b9992075c7127ff4e2b5361dd57b732f86b97ad
test case for debugger stmt in reactive scope
[ { "path": "compiler/forget/src/__tests__/fixtures/compiler/debugger-memoized.expect.md", "patch": "@@ -0,0 +1,36 @@\n+\n+## Input\n+\n+```javascript\n+function Component(props) {\n+ const x = [];\n+ debugger;\n+ x.push(props.value);\n+ return x;\n+}\n+\n+```\n+\n+## Code\n+\n+```javascript\n+import { un...
2023-04-26T18:34:46
huggingface/transformers
372c27e71f80e64571ac1149d1708e641d7d44da
8efd81513f51cd343d086e0e8f69190caa49dbcf
Train sampler unification (#43138) * add sequential sampling * train_sampler arg * Update src/transformers/training_args.py Co-authored-by: Marc Sun <57196510+SunMarc@users.noreply.github.com> * updated train_sampler to train_sampling_strategy; updated warnings/docs * update docs that prev reference group_by_leng...
[ { "path": "docs/source/en/model_doc/exaone_moe.md", "patch": "@@ -13,7 +13,7 @@ specific language governing permissions and limitations under the License.\n rendered properly in your Markdown viewer.\n \n -->\n-*This model was released on 2025-12-31 and added to Hugging Face Transformers on 2026-02-02.*\n+*...
2026-02-11T10:38:35
golang/go
00709919d09904b17cfe3bfeb35521cbd3fb04f8
4865aadc21acebc8039f914929f03c7042b2ae8d
cmd/compile, cmd/link: FIPS fixes for large programs 1. In cmd/internal/obj, only apply the exclusion list to data symbols. Text symbols are always fine since they can use PC-relative relocations. 2. In cmd/link, only skip trampolines for text symbols in the same package with the same type. Before, all text sym...
[ { "path": "src/cmd/internal/obj/fips140.go", "patch": "@@ -238,27 +238,30 @@ func (s *LSym) setFIPSType(ctxt *Link) {\n \n \t// Now we're at least handling a FIPS symbol.\n \t// It's okay to be slower now, since this code only runs when compiling a few packages.\n+\t// Text symbols are always okay, since th...
2024-11-22T01:03:18
ggml-org/llama.cpp
0eb4764182df1f031f92b4121f3bf8fbe8026565
1f5d15e6657c119251e2a2c2c69333660f8cb553
vulkan: add noncontiguous GLU support (#21081) * vulkan: add noncontiguous GLU support * fix compile issue
[ { "path": "ggml/src/ggml-vulkan/ggml-vulkan.cpp", "patch": "@@ -1112,6 +1112,16 @@ struct vk_op_glu_push_constants {\n uint32_t mode; // 0: default, 1: swapped, 2: split\n float alpha; // for swiglu_oai\n float limit;\n+ uint32_t nb01;\n+ uint32_t nb02;\n+ uint32_t nb03;\n+ uint32_t...
2026-03-28T07:44:56
electron/electron
652680e8012983b1d8b9118b48297df4c1d9a939
cf3ee7be56dad64858a243581eac0644d4653488
fix: ensure external memory adjustments are balanced (#33266)
[ { "path": "shell/common/api/electron_api_native_image.cc", "patch": "@@ -109,7 +109,7 @@ base::win::ScopedHICON ReadICOFromPath(int size, const base::FilePath& path) {\n \n NativeImage::NativeImage(v8::Isolate* isolate, const gfx::Image& image)\n : image_(image), isolate_(isolate) {\n- AdjustAmountOfEx...
2022-03-16T17:54:30
facebook/react
8e7b68506f0b490c7eca3ee886bfd1ff7991d07c
1f51b10ac9762550574485028a4081902d6bd034
You get a debugger and you get a debugger and you get a Adds support for DebuggerStatement.
[ { "path": "compiler/forget/src/HIR/BuildHIR.ts", "patch": "@@ -866,9 +866,22 @@ function lowerStatement(\n );\n return;\n }\n+ case \"DebuggerStatement\": {\n+ const stmt = stmtPath as NodePath<t.DebuggerStatement>;\n+ const loc = stmt.node.loc ?? GeneratedSource;\n+ builde...
2023-04-26T18:27:37
huggingface/transformers
8efd81513f51cd343d086e0e8f69190caa49dbcf
6c4710bdef6b0db127d40704d8ee2f82c551b27f
revert place_model_on_device to property (#43895) * revert place_model_on_device to property * Apply repo consistency fixes --------- Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
[ { "path": "src/transformers/trainer.py", "patch": "@@ -340,7 +340,7 @@ class Trainer:\n data parallelism, this means some of the model layers are split on different GPUs).\n - **place_model_on_device** -- Whether or not to automatically place the model on the device. Defaults to\n ...
2026-02-11T10:37:04
ggml-org/llama.cpp
1f5d15e6657c119251e2a2c2c69333660f8cb553
c46758d28fa9846893f37e8cec03b73fee120604
common/parser: fix reasoning whitespace bugs + extra parser tests (#21085) * fix whitespace reasoning issues + add reconstruction tests * Proper fix * fix Nemotron autoparser test expectations to include newline in marker
[ { "path": "common/chat-diff-analyzer.cpp", "patch": "@@ -287,7 +287,7 @@ void analyze_reasoning::compare_reasoning_presence() {\n return p.literal(reasoning_content) + p.space() + p.optional(p.tag(\"post\", (p.marker() + p.space())) + p.rest());\n });\n auto parser_wrapped = buil...
2026-03-28T06:29:26
golang/go
4865aadc21acebc8039f914929f03c7042b2ae8d
de76c0dff7b43a761786a9b5c75db2fb77797d62
cmd/dist: don't test FIPS when ASAN is on CL 627603 added a clear error that FIPS+ASAN doesn't work, and disabled a test in check_test.go. The :gofips140 test variants in cmd/dist need to be disabled as well. Remove a return after testing.T.Skipf since it's unreachable. For #70321. Fixes #70496. Change-Id: Ia53830d...
[ { "path": "src/cmd/dist/test.go", "patch": "@@ -714,7 +714,7 @@ func (t *tester) registerTests() {\n \t})\n \n \t// Check that all crypto packages compile (and test correctly, in longmode) with fips.\n-\tif fipsSupported() {\n+\tif t.fipsSupported() {\n \t\t// Test standard crypto packages with fips140=on.\...
2024-11-22T04:39:47
electron/electron
b27401172033e7c8fac38d3b2d784fe8f9a481b2
02fe2455216578815d6300ba2a44d462a840142d
fix: fiddle ipc code pattern 3 (#33262)
[ { "path": "docs/fiddles/ipc/pattern-3/renderer.js", "patch": "@@ -4,5 +4,5 @@ window.electronAPI.handleCounter((event, value) => {\n const oldValue = Number(counter.innerText)\n const newValue = oldValue + value\n counter.innerText = newValue\n- event.reply('counter-value', newValue)\n+ ev...
2022-03-16T15:01:29
facebook/react
1f51b10ac9762550574485028a4081902d6bd034
2cb1c723051378a3e94bec445f135a457d93f04f
Fix unused logical/condition via ExpressionStatement instr Uses the new ExpressionStatement instruction to ensure that logical and conditional expressions are never pruned. This addresses an issue where we were unable to construct a ReactiveFunction for unused logical/conditional bc there wasn't a single Identifie...
[ { "path": "compiler/forget/src/HIR/BuildHIR.ts", "patch": "@@ -674,7 +674,24 @@ function lowerStatement(\n case \"ExpressionStatement\": {\n const stmt = stmtPath as NodePath<t.ExpressionStatement>;\n const expression = stmt.get(\"expression\");\n- lowerExpressionToTemporary(builder, ex...
2023-04-26T18:27:36
huggingface/transformers
42dc276ca893df70ae3265a7e4fa13a31697e5a6
e2372ffabdf6da99364d3d053da7783acf5f623c
Fix typo in quantization_operations in PEFT integrations (#43821) fix typo
[ { "path": "src/transformers/integrations/peft.py", "patch": "@@ -261,7 +261,7 @@ def _build_peft_weight_mapping(\n source_patterns=new_source_patterns,\n target_patterns=new_target_patterns,\n distributed_operation=orig_conversion.distributed_opera...
2026-02-11T10:09:22
ggml-org/llama.cpp
c46758d28fa9846893f37e8cec03b73fee120604
bf934f28db9da38d1ecfdd0380a94170c34f42ff
cli : add /glob command (#21084) * add /glob command * output error when max files reached * support globbing outside curdir
[ { "path": "common/common.cpp", "patch": "@@ -656,6 +656,38 @@ bool string_parse_kv_override(const char * data, std::vector<llama_model_kv_over\n return true;\n }\n \n+// simple glob: * matches non-/ chars, ** matches anything including /\n+static inline bool glob_match(const char * pattern, const char *...
2026-03-28T01:33:04
golang/go
4b7f7cd87dfcbc17861c908b20a6101e5915ef59
dbfd0036f7c25eef6c194bd37936398ce58f910d
crypto/tls: implement X25519MLKEM768 This makes three related changes that work particularly well together and would require significant extra work to do separately: it replaces X25519Kyber768Draft00 with X25519MLKEM768, it makes CurvePreferences ordering crypto/tls-selected, and applies a preference to PQ key exchang...
[ { "path": "api/next/69985.txt", "patch": "@@ -0,0 +1,2 @@\n+pkg crypto/tls, const X25519MLKEM768 = 4588 #69985\n+pkg crypto/tls, const X25519MLKEM768 CurveID #69985", "additions": 2, "deletions": 0, "language": "Plain Text" }, { "path": "doc/godebug.md", "patch": "@@ -203,6 +203,11 @...
2024-11-21T20:21:58
ggml-org/llama.cpp
bf934f28db9da38d1ecfdd0380a94170c34f42ff
5c1a7b83550b4f1e0954007556f9b28f7db1a427
docker : fix and enable ARM64 image build (#20929) * CI: fix ARM64 image build error & enable compilation * Update .github/workflows/docker.yml Co-authored-by: Aaron Teo <taronaeo@gmail.com> * CI: revert ggml/src/ggml-cpu/CMakeLists.txt * Update .github/workflows/docker.yml Co-authored-by: Aaron Teo <taronaeo@gma...
[ { "path": ".devops/cpu.Dockerfile", "patch": "@@ -1,11 +1,13 @@\n-ARG UBUNTU_VERSION=22.04\n+ARG UBUNTU_VERSION=24.04\n \n FROM ubuntu:$UBUNTU_VERSION AS build\n \n ARG TARGETARCH\n \n RUN apt-get update && \\\n- apt-get install -y build-essential git cmake libssl-dev\n+ apt-get install -y gcc-14 g++-...
2026-03-28T00:45:09
huggingface/transformers
cc88192bc5ca0b66e7fc1152bb372fb90c6bc2ce
e00bac44245f93f0f93c1440c8f161d5551bb089
Fix expected DAC outputs due to (old) change in CI settings. (#43896) * Fix expected DAC expected outputs. * Add reference to PR that caused change
[ { "path": "tests/models/dac/test_modeling_dac.py", "patch": "@@ -307,6 +307,7 @@ def compute_rmse(arr1, arr2):\n Code for reproducing expected outputs can be found here:\n - test_integration: https://gist.github.com/ebezzam/bb315efa7a416db6336a6b2a2d424ffa#file-test_dac-py\n - test_batch: https://gist.githu...
2026-02-10T17:47:59
ollama/ollama
40b10eee6d62a32578ca7e884fb73d4c8bc644a0
424f648632c925ce14a75018c4dcab395e035993
api: fix ImageData struct comment to expect raw image bytes (#10386)
[ { "path": "api/types.go", "patch": "@@ -76,7 +76,7 @@ type GenerateRequest struct {\n \t// this request.\n \tKeepAlive *Duration `json:\"keep_alive,omitempty\"`\n \n-\t// Images is an optional list of base64-encoded images accompanying this\n+\t// Images is an optional list of raw image bytes accompanying t...
2025-04-24T03:13:51
golang/go
dbfd0036f7c25eef6c194bd37936398ce58f910d
b299e9a44f298e72815ca0513bcc6ccca075f3fc
net/http: re-enable TestTransportServerProtocols The h2_bundle.go update was done in CL 631035, and the test now passes. Fixes #67816. Change-Id: Icd54c2774a9e2781e7dc9342ae70c3034eb9bab9 Reviewed-on: https://go-review.googlesource.com/c/go/+/631037 Auto-Submit: Damien Neil <dneil@google.com> Auto-Submit: Dmitri Shu...
[ { "path": "src/net/http/transport_test.go", "patch": "@@ -7173,8 +7173,6 @@ func testValidateClientRequestTrailers(t *testing.T, mode testMode) {\n }\n \n func TestTransportServerProtocols(t *testing.T) {\n-\tt.Skip(\"https://go.dev/issue/67816: skip test until h2_bundle.go is updated\")\n-\n \tCondSkipHTTP...
2024-11-22T02:24:11
electron/electron
c8a3a00017d5eaca4e017a22d63acaebeb1e24d4
73828910152cb67c849635ccf513dbe107b63ca3
fix: race condition where webContents can be nullptr during re-focus and a multi-window close sequence (#33063) * fix: race condition where webContents can be nullptr during re-focus and a multi-window close sequence * chore: update electron_inspectable_web_contents_view.mm
[ { "path": "shell/browser/ui/cocoa/electron_inspectable_web_contents_view.mm", "patch": "@@ -266,6 +266,8 @@ - (void)viewDidBecomeFirstResponder:(NSNotification*)notification {\n inspectableWebContentsView_->inspectable_web_contents();\n DCHECK(inspectable_web_contents);\n auto* webContents = inspe...
2022-03-15T17:45:56
ggml-org/llama.cpp
5c1a7b83550b4f1e0954007556f9b28f7db1a427
59d840209a5195c2f6e2e81b5f8339a0637b59d9
server : add custom socket options to disable SO_REUSEPORT (#21056) * server : add custom socket options to disable SO_REUSEPORT Signed-off-by: Adrien Gallouët <angt@huggingface.co> * Add --reuse-port $ strace -e trace=setsockopt,bind build/bin/llama-server -lv 2 --reuse-port setsockopt(3, SOL_TCP, TCP_NODE...
[ { "path": "common/arg.cpp", "patch": "@@ -2807,6 +2807,13 @@ common_params_context common_params_parser_init(common_params & params, llama_ex\n params.port = value;\n }\n ).set_examples({LLAMA_EXAMPLE_SERVER}).set_env(\"LLAMA_ARG_PORT\"));\n+ add_opt(common_arg(\n+ {\"--reu...
2026-03-28T00:12:43
golang/go
07b42666051841352077c0d04ba67d510247fd1d
9aaef91d3e75994f19d33a4fcee08b5225432af6
crypto/x509: generate serial number for nil template SerialNumber Fixes #67675 Change-Id: I976935d20eb6b9adcd19d47bcaeb7abcf78ec5bb Reviewed-on: https://go-review.googlesource.com/c/go/+/630995 Reviewed-by: Roland Shoemaker <roland@golang.org> LUCI-TryBot-Result: Go LUCI <golang-scoped@luci-project-accounts.iam.gserv...
[ { "path": "src/crypto/x509/x509.go", "patch": "@@ -27,6 +27,7 @@ import (\n \t\"crypto/ecdsa\"\n \t\"crypto/ed25519\"\n \t\"crypto/elliptic\"\n+\tcryptorand \"crypto/rand\"\n \t\"crypto/rsa\"\n \t\"crypto/sha1\"\n \t\"crypto/x509/pkix\"\n@@ -1655,6 +1656,9 @@ var emptyASN1Subject = []byte{0x30, 0}\n // If S...
2024-11-22T00:24:09
ollama/ollama
424f648632c925ce14a75018c4dcab395e035993
2eb1fb3231063365408155d2fffce9d62ad3c5ee
increase default context length to 4096 (#10364) * increase default context length to 4096 We lower the default numParallel from 4 to 2 and use these "savings" to double the default context length from 2048 to 4096. We're memory neutral in cases when we previously would've used numParallel == 4, but we add the follo...
[ { "path": "cmd/cmd.go", "patch": "@@ -1407,6 +1407,7 @@ func NewCLI() *cobra.Command {\n \t\t\t\tenvVars[\"OLLAMA_LLM_LIBRARY\"],\n \t\t\t\tenvVars[\"OLLAMA_GPU_OVERHEAD\"],\n \t\t\t\tenvVars[\"OLLAMA_LOAD_TIMEOUT\"],\n+\t\t\t\tenvVars[\"OLLAMA_CONTEXT_LENGTH\"],\n \t\t\t})\n \t\tdefault:\n \t\t\tappendEnvD...
2025-04-22T23:33:24
huggingface/transformers
42791a34fdeae197f60f11ace3807c81f44b0729
3d24ffaa4f0927cc8b5da10157ac009d34942472
Separate `check_model_inputs` into `capture_outputs` and `merge_with_config_defaults` + ensure correctness (#43862) * start * switch imports everywhere * simplify * fix * latest model * add test * improve test * fix * fix * fix * oupsi * doc * better test * fix a few * fix * more elegant fix * finalize...
[ { "path": "examples/modular-transformers/modeling_dummy_bert.py", "patch": "@@ -19,7 +19,8 @@\n from ...processing_utils import Unpack\n from ...pytorch_utils import apply_chunking_to_forward\n from ...utils import TransformersKwargs, auto_docstring\n-from ...utils.generic import check_model_inputs\n+from ....
2026-02-10T17:37:41
golang/go
918765b619b87a603c40b331bd5f5e2335d265f0
fb432782f0017f59f6f2c5f620dfd8c67b5e35e3
crypto/x509: switch default policy field to Policies Switch from Certificate.PolicyIdentifiers to Certificate.Policies when marshalling. Fixes #67620 Change-Id: Ib627135a569f53d344b4ee2f892ba139506ce0d1 Reviewed-on: https://go-review.googlesource.com/c/go/+/629855 Reviewed-by: Damien Neil <dneil@google.com> Reviewed...
[ { "path": "doc/godebug.md", "patch": "@@ -195,6 +195,14 @@ This currently only affects arm64 programs. For all other platforms it is a no-o\n Go 1.24 removed the `x509sha1` setting. `crypto/x509` no longer supports verifying\n signatures on certificates that use SHA-1 based signature algorithms.\n \n+Go 1....
2024-11-19T22:05:38
ollama/ollama
4e535e618846ffb00a2a6714c07847d6d2951453
40b8fdbdcacb41b9cf42869051df765f66750036
server/internal/registry: make pull send errors with Error field (#10326) Previously, the pull handler would send an error message in the Status field, this prevented the client from using the message as a signal to stop. In the case of the "run" command, it would follow the pull with a "show" which would print a near...
[ { "path": "server/internal/registry/server.go", "patch": "@@ -244,6 +244,7 @@ func (s *Local) handleDelete(_ http.ResponseWriter, r *http.Request) error {\n }\n \n type progressUpdateJSON struct {\n+\tError string `json:\"error,omitempty,omitzero\"`\n \tStatus string `json:\"status,omitempt...
2025-04-19T01:12:28
ggml-org/llama.cpp
59d840209a5195c2f6e2e81b5f8339a0637b59d9
ff934e29bc1d102bb77418f4c83a79d954220753
common : inhibit lazy grammar sampler while reasoning is active (#20970) * common : inhibit grammar while reasoning budget is active * cont : update force_pos in accept * cont : fix tests * cont : tweak should apply logic * cont : return early not using grammar sampler * Add tests * cont : prevent backend sampli...
[ { "path": "common/reasoning-budget.cpp", "patch": "@@ -115,9 +115,11 @@ static void common_reasoning_budget_accept(struct llama_sampler * smpl, llama_to\n break;\n }\n case REASONING_BUDGET_FORCING:\n- // force_pos is advanced in apply(), not here.\n- // Thi...
2026-03-27T17:30:40
huggingface/transformers
3d24ffaa4f0927cc8b5da10157ac009d34942472
15bada75ffe82895a6fe8eb6b7096c0c6b7e9b33
[fix] DAC model: Apply STE in Dac.from_latents to match the forward pass (#43820) * fix(models): Apply STE in DacResidualVectorQuantizer.from_latents * nit: Add quantizer-level assert to verify matching latents --------- Co-authored-by: Anton Vlasjuk <73884904+vasqu@users.noreply.github.com> Co-authored-by: Eric Be...
[ { "path": "src/transformers/models/dac/modeling_dac.py", "patch": "@@ -390,11 +390,13 @@ def from_latents(self, latents: torch.Tensor):\n n_codebooks = np.where(dims <= latents.shape[1])[0].max(axis=0, keepdims=True)[0]\n for i in range(n_codebooks):\n hidden_dim_j, hidden_dim_k ...
2026-02-10T17:15:20
facebook/react
894cf6e37b7b34c95076328a10efbbc2117f89bc
9aea37ff5af310dbd28a22b0d56e3d4233e8cb24
First-class representation of builtin jsx tags We previously represented JsxExpressions using builtin tags - `<div>`, `<b>` etc - by lowering the tag name to a Primitive with the string name of the tag. However, by lowering into an independent value, it was possible that the lowered tag name could be grouped into ...
[ { "path": "compiler/forget/src/HIR/BuildHIR.ts", "patch": "@@ -17,6 +17,7 @@ import {\n ArrayPattern,\n BlockId,\n BranchTerminal,\n+ BuiltinTag,\n Case,\n Effect,\n GeneratedSource,\n@@ -2035,7 +2036,7 @@ function lowerJsxElementName(\n exprPath: NodePath<\n t.JSXIdentifier | t.JSXMember...
2023-04-26T18:27:35
golang/go
8b97607280810d8effcfda56e47aa3ff73dced67
8cecfad2a99987a35edfbcd875bef5e894abbce7
crypto/sha3: new package Implement the SHA-3 hash algorithms and the SHAKE extendable output functions defined in FIPS 202. This is a wrapper for crypto/internal/fips/sha3 which in turn was ported from x/crypto/sha3 in CL 616717 as part of #65269. Fixes #69982 Change-Id: I64ce7f362c1a773f7f5b05f7e0acb4110e52a329 Re...
[ { "path": "api/next/69982.txt", "patch": "@@ -0,0 +1,31 @@\n+pkg crypto/sha3, func New224() *SHA3 #69982\n+pkg crypto/sha3, func New256() *SHA3 #69982\n+pkg crypto/sha3, func New384() *SHA3 #69982\n+pkg crypto/sha3, func New512() *SHA3 #69982\n+pkg crypto/sha3, func NewCSHAKE128([]uint8, []uint8) *SHAKE #69...
2024-11-18T15:58:49
ggml-org/llama.cpp
ee051c1e4e6ceddc2fa516eb067496328ac1a2dd
e6f6770515d4b5db2300f32ec6bb504e997774df
hexagon: support for IQ4_NL and MXFP4 (#21018) * ggml-hexagon: add IQ4_NL and MXFP4 HMX matmul support - Add IQ4_NL quantization type support to Hexagon backend (buffer set/get tensor repack, mul_mat, mul_mat_id dispatch) - Implement HVX IQ4_NL vec_dot kernels (1x1, 2x1, 2x2) with LUT-based 4-bit index to int8 kv...
[ { "path": "ggml/src/ggml-hexagon/ggml-hexagon.cpp", "patch": "@@ -1406,6 +1406,13 @@ static void ggml_backend_hexagon_buffer_set_tensor(ggml_backend_buffer_t buffer,\n repack_q8_0_q8x4x2(tensor, data, size);\n break;\n \n+ case GGML_TYPE_IQ4_NL:\n+ GGML_ASSERT(offse...
2026-03-27T16:22:41
ollama/ollama
1d99451ad705478c0a22262ad38b5a403b61c291
09bb2e30f69489b2bd5138fa81d9dbb54c1d2f19
server/internal/client/ollama: handle some network errors gracefully (#10317)
[ { "path": "server/internal/client/ollama/registry.go", "patch": "@@ -223,8 +223,21 @@ type Registry struct {\n \tChunkingThreshold int64\n \n \t// Mask, if set, is the name used to convert non-fully qualified names\n-\t// to fully qualified names. If empty, [DefaultMask] is used.\n+\t// to fully qualified n...
2025-04-17T19:43:09
huggingface/transformers
60a5dbaaf52fa0273ed7a1d55536112b24ee2676
71d7fc58af53f6fd22cd9a398ca618905abf0d9e
[Docs] Add docs for GLM-OCR and fix EomT-DINOv3 (#43710) * Update docs * Remove CLAUDE.md * Apply suggestions from code review Co-authored-by: Steven Liu <59462357+stevhliu@users.noreply.github.com> * Fix missing comma in GLM-OCR docs flash attention example Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>...
[ { "path": "docs/source/en/model_doc/eomt_dinov3.md", "patch": "@@ -43,7 +43,7 @@ Tips:\n combined with the upstream DINOv3 backbones. The conversion utilities in the\n [official repository](https://github.com/tue-mps/eomt) describe this workflow in detail.\n \n-This model was contributed by [NielsRogge]...
2026-02-10T15:35:22
ggml-org/llama.cpp
ba38f3becce7d1283585c73d796eb47d72bbbd30
37f230dd7cff6732580bec976ca9e378af230d20
rpc : proper handling of data pointers to CPU buffers (#21030) The compute graph may contain tensors pointing to CPU buffers. In these cases the buffer address is serialized as 0 and sent over the wire. However, the data pointer is serialized as-is and this prevents proper validation on the server side. This patches f...
[ { "path": "ggml/src/ggml-rpc/ggml-rpc.cpp", "patch": "@@ -589,8 +589,10 @@ static rpc_tensor serialize_tensor(const ggml_tensor * tensor) {\n ggml_backend_buffer_t buffer = tensor->buffer;\n ggml_backend_rpc_buffer_context * ctx = (ggml_backend_rpc_buffer_context *)buffer->context;\n ...
2026-03-27T08:59:35
electron/electron
039c061d07821b2d77781320a0bdf60d2e1b938a
f37295325668e50c80fbc00b406f0b4660cc93cd
fix: add missing [[maybe_unused]] to IsEnvSet (#33224)
[ { "path": "shell/app/electron_main_mac.cc", "patch": "@@ -20,7 +20,7 @@\n \n namespace {\n \n-bool IsEnvSet(const char* name) {\n+[[maybe_unused]] bool IsEnvSet(const char* name) {\n char* indicator = getenv(name);\n return indicator && indicator[0] != '\\0';\n }", "additions": 1, "deletions": 1...
2022-03-14T09:19:15
golang/go
8cecfad2a99987a35edfbcd875bef5e894abbce7
3b42687c56af9a7e52a8194e75c0dc14962b37b8
crypto/rsa: port Validate to bigmod This is quite a bit slower (almost entirely in the e * d reductions, which could be optimized), but the slowdown is only 12% of a signature operation. Also, call Validate at the end of GenerateKey as a backstop. Key generation is so incredibly slow that the extra time is negligible...
[ { "path": "src/crypto/internal/fips140/bigmod/nat.go", "patch": "@@ -202,6 +202,19 @@ func (x *Nat) setBytes(b []byte) error {\n \treturn nil\n }\n \n+// SetUint assigns x = y, and returns an error if y >= m.\n+//\n+// The output will be resized to the size of m and overwritten.\n+func (x *Nat) SetUint(y ui...
2024-11-21T12:51:21
ollama/ollama
3457a315b241d5d2ada9958d22cc5effb2643a7e
ed4e1393149e1ba5e8fbf5b6629d2658342e39d9
server/internal/client/ollama: cleanup use of multiple counters (#10304) The completed and received counters must work in tandem and the code should better reflect that. Previously, the act of updating them was 2-3 lines of code duplicated in multiple places. This consolidates them into a single update closure for eas...
[ { "path": "server/internal/client/ollama/registry.go", "patch": "@@ -431,14 +431,13 @@ func (r *Registry) Push(ctx context.Context, name string, p *PushParams) error {\n //\n // It always calls update with a nil error.\n type trackingReader struct {\n-\tl *Layer\n \tr io.Reader\n-\tupdate func(n i...
2025-04-16T21:33:40
huggingface/transformers
71d7fc58af53f6fd22cd9a398ca618905abf0d9e
e12aa2d22b0088fa00c697f7d8987a744fe9da7b
[`fix`] Use `last_hidden_state` key from `get_image_features` for llama4 (#43882) Use last_hidden_state key from get_image_features for llama4
[ { "path": "src/transformers/models/llama4/modeling_llama4.py", "patch": "@@ -1080,7 +1080,7 @@ def forward(\n output_hidden_states: bool | None = None,\n return_dict: bool | None = None,\n **kwargs,\n- ) -> BaseModelOutput | tuple[torch.Tensor, ...]:\n+ ) -> BaseModelOutputWith...
2026-02-10T15:05:50
ggml-org/llama.cpp
a308e584cae3fa8cee1d739a858a2d780f1de009
d0fa2c9fbb8414a509fc8c4f534ad3c03e591735
completion : Fix segfault on model load failure (#21049)
[ { "path": "tools/completion/completion.cpp", "patch": "@@ -146,13 +146,19 @@ int main(int argc, char ** argv) {\n \n ctx = llama_init->context();\n model = llama_init->model();\n- smpl = llama_init->sampler(0);\n \n if (ctx == NULL) {\n LOG_ERR(\"%s: error: unable to create context...
2026-03-27T08:01:13
facebook/react
5f2f76cbcd27eb1bd2d0cb68965046183f71d98b
5ed297fa74fb75d4b25d8b3e9dfa0608c23872aa
Fix test262 We [renamed the export](https://github.com/facebook/react-forget/blob/6f001f54b036d6f427c8d4c1c04ace3ab1998a4d/forget/src/Babel/RunReactForgetBabelPlugin.ts#L22) of runReactForgetBabelPlugin so the test262 preprocessor was erroring
[ { "path": "compiler/forget/scripts/test262-preprocessor.js", "patch": "@@ -5,8 +5,9 @@\n * LICENSE file in the root directory of this source tree.\n */\n \n-const runReactForgetBabelPlugin =\n- require(\"../dist/Babel/RunReactForgetBabelPlugin\").default;\n+const {\n+ runReactForgetBabelPlugin,\n+} = re...
2023-04-26T18:12:18
electron/electron
f37295325668e50c80fbc00b406f0b4660cc93cd
cdc27a3793958de4a2c06e7fce16244b6f0cb354
fix: cppgc/node collisions in renderer process (#33252) * fix: cppgc/node collisions in renderer process * Update be_compatible_with_cppgc.patch
[ { "path": "patches/node/.patches", "patch": "@@ -32,3 +32,4 @@ darwin_remove_eprototype_error_workaround_3405.patch\n darwin_translate_eprototype_to_econnreset_3413.patch\n darwin_bump_minimum_supported_version_to_10_15_3406.patch\n fix_failing_node_js_test_on_outdated.patch\n+be_compatible_with_cppgc.patch...
2022-03-12T23:28:10
golang/go
10fb001c7519f2a694e987ecb1e75da80f832b41
ca14eaf77c86bd5492329d2be6f1a82afe7802f5
crypto/rsa: refuse to generate and/or use keys smaller than 1024 bits Fixes #68762 Change-Id: Id89c770571d7cc27c6cf7932139ec3424383a7ef Reviewed-on: https://go-review.googlesource.com/c/go/+/629938 Reviewed-by: Roland Shoemaker <roland@golang.org> Auto-Submit: Filippo Valsorda <filippo@golang.org> Reviewed-by: Dmitri...
[ { "path": "doc/godebug.md", "patch": "@@ -177,6 +177,10 @@ This behavior can be controlled with the `gotestjsonbuildtext` setting.\n Using `gotestjsonbuildtext=1` restores the 1.23 behavior.\n This setting will be removed in a future release, Go 1.28 at the earliest.\n \n+Go 1.24 changed [`crypto/rsa`](/pkg...
2024-11-20T12:59:09
ollama/ollama
56dc316a57f07fbed80723d1ecd589da0906018e
2fec73eef6e9482f606f185ebb2ae4f75ad1a37c
Give tests more time to run (#10306) Fix flake failures on windows
[ { "path": "server/sched_test.go", "patch": "@@ -355,7 +355,7 @@ func TestRequestsMultipleLoadedModels(t *testing.T) {\n }\n \n func TestGetRunner(t *testing.T) {\n-\tctx, done := context.WithTimeout(context.Background(), 200*time.Millisecond)\n+\tctx, done := context.WithTimeout(context.Background(), 3*time...
2025-04-16T20:37:00
huggingface/transformers
e12aa2d22b0088fa00c697f7d8987a744fe9da7b
6d86e9a318f67f936f672e489129d455852eed99
Refactor trainer init (#43807) * refactor trainer init * update init * simplify liger * udapte * better * comments * do_train not reliable at all * This should make more sense now * Apply suggestion from @qgallouedec Co-authored-by: Quentin Gallouédec <45557362+qgallouedec@users.noreply.github.com> * Apply s...
[ { "path": "src/transformers/integrations/__init__.py", "patch": "@@ -116,6 +116,7 @@\n \"run_hp_search_ray\",\n \"run_hp_search_wandb\",\n ],\n+ \"liger\": [\"apply_liger_kernel\"],\n \"moe\": [\n \"batched_mm_experts_forward\",\n \"grouped_mm_experts_forward\",\n@...
2026-02-10T15:00:18
ggml-org/llama.cpp
9bcb4eff4d7e7755276645c767948c25c00481b8
6861f6509a2b5b3b1b21d04fc502ba55f1326a7d
metal : Fix dimension constraint violation in matmul2d descriptor (#21048) Updates Metal tensor API test probe to fix the dimension constraint violation in the matmul2d descriptor (at least one value must be a multiple of 16).
[ { "path": "ggml/src/ggml-metal/ggml-metal-device.m", "patch": "@@ -690,7 +690,7 @@ ggml_metal_device_t ggml_metal_device_init(int device) {\n \" auto tB = B.slice((int)tgid.x, 0); \\n\"\n \" \\n\"\n \" matmul2d< \\n\"\n- \"...
2026-03-27T07:05:21
facebook/react
6eadbe0c4aebf68410bb48147054ee22eec4c20c
ec5e9c2a75749b0a470b7148738cb85bbb035958
Fix: Resolve entangled actions independently (#26726) When there are multiple async actions at the same time, we entangle them together because we can't be sure which action an update might be associated with. (For this, we'd need AsyncContext.) However, if one of the async actions fails with an error, it should on...
[ { "path": "packages/react-reconciler/src/ReactFiberAsyncAction.js", "patch": "@@ -7,44 +7,36 @@\n * @flow\n */\n \n-import type {Wakeable} from 'shared/ReactTypes';\n+import type {\n+ Thenable,\n+ PendingThenable,\n+ FulfilledThenable,\n+ RejectedThenable,\n+} from 'shared/ReactTypes';\n import type {...
2023-04-26T00:43:20
electron/electron
cdc27a3793958de4a2c06e7fce16244b6f0cb354
bbb79880f7016c156c8fad3c08b59caa0eedd8a7
fix: prevent UAF crash in setCertificateVerifyProc (#33204)
[ { "path": "patches/chromium/expose_setuseragent_on_networkcontext.patch", "patch": "@@ -33,10 +33,10 @@ index 14c71cc69388da46f62d9835e2a06fef0870da02..9481ea08401ae29ae9c1d960491b05b3\n \n } // namespace net\n diff --git a/services/network/network_context.cc b/services/network/network_context.cc\n-index...
2022-03-11T19:35:48
ollama/ollama
2fec73eef6e9482f606f185ebb2ae4f75ad1a37c
1e7f62cb429e5a962dd9c448e7b1b3371879e48b
fix write gguf padding
[ { "path": "convert/convert_llama.go", "patch": "@@ -28,12 +28,12 @@ type llamaModel struct {\n \tNumKeyValueHeads uint32 `json:\"num_key_value_heads\"`\n \tRopeTheta float32 `json:\"rope_theta\"`\n \tRopeScaling struct {\n-\t\tType string `json:\"type\...
2025-04-11T20:39:51
huggingface/transformers
6d86e9a318f67f936f672e489129d455852eed99
b33235e3c03a96ef5fc2cc084e368b658062ffa7
fix fbgemm fp8 multi-device load failure. (#43581) Signed-off-by: Wang, Yi <yi.a.wang@intel.com> Co-authored-by: Marc Sun <57196510+SunMarc@users.noreply.github.com> Co-authored-by: Mohamed Mekkouri <93391238+MekkCyber@users.noreply.github.com>
[ { "path": "src/transformers/quantizers/quantizer_fbgemm_fp8.py", "patch": "@@ -108,6 +108,13 @@ def param_needs_quantization(self, model: \"PreTrainedModel\", param_name: str, **\n return True\n return False\n \n+ def param_element_size(self, model: \"PreTrainedModel\", param_name...
2026-02-10T14:51:48
ggml-org/llama.cpp
6861f6509a2b5b3b1b21d04fc502ba55f1326a7d
1743d9805706fba86e89bff1de0b6410dd56252d
CANN: update docker images to 8.5.0 and improve CANN.md (#20801) * cann: update docker images to 8.5.0 - bump CANN base image from 8.3.rc2 to 8.5.0 - bump ASCEND_VERSION from 8.1.RC1.alpha001 to 8.5.0 Move to newer stable releases. * cann: update CANN.md * Update CANN.md to include BF16 support Added BF16 support...
[ { "path": ".devops/cann.Dockerfile", "patch": "@@ -4,7 +4,7 @@\n \n # Define the CANN base image for easier version updates later\n ARG CHIP_TYPE=910b\n-ARG CANN_BASE_IMAGE=quay.io/ascend/cann:8.3.rc2-${CHIP_TYPE}-openeuler24.03-py3.11\n+ARG CANN_BASE_IMAGE=quay.io/ascend/cann:8.5.0-${CHIP_TYPE}-openeuler24...
2026-03-27T00:53:00
facebook/react
ec5e9c2a75749b0a470b7148738cb85bbb035958
f87e97a0a67fa7cfd7e6f2ec985621c0e825cb23
Fix double preload (#26729) I found a couple scenarios where preloads were issued too aggressively 1. During SSR, if you render a new stylesheet after the preamble flushed it will flush a preload even if the resource was already preloaded 2. During Client render, if you call `ReactDOM.preload()` it will only che...
[ { "path": "packages/react-dom-bindings/src/client/ReactFiberConfigDOM.js", "patch": "@@ -2138,8 +2138,12 @@ function preload(href: string, options: PreloadOptions) {\n const as = options.as;\n const limitedEscapedHref =\n escapeSelectorAttributeValueInsideDoubleQuotes(href);\n- const preloa...
2023-04-25T22:10:34
electron/electron
bbb79880f7016c156c8fad3c08b59caa0eedd8a7
dc63b8e7f478ad87bcde0406508c1fbbcf4b872c
chore: cherry-pick 2ed58f4 from chromium (#33109) Refs: https://chromium-review.googlesource.com/c/chromium/src/+/3492658 Fixes: https://github.com/electron/electron/issues/33049 Signed-off-by: Darshan Sen <raisinten@gmail.com> Co-authored-by: Charles Kerr <charles@charleskerr.com>
[ { "path": "patches/chromium/.patches", "patch": "@@ -114,3 +114,4 @@ port_autofill_colors_to_the_color_pipeline.patch\n build_disable_partition_alloc_on_mac.patch\n build_disable_thin_lto_on_mac.patch\n fix_non-client_mouse_tracking_and_message_bubbling_on_windows.patch\n+remove_incorrect_width_height_adjus...
2022-03-11T17:19:51
ollama/ollama
ccb7eb81357c6612705c6b67306616e2bc57acf7
0fe487e7325cd72b8b06e758c2a03842c557dbda
ggml: Free ggml_backend_buffer_t when releasing buffer When ggml_backend_buffer_free() is called, the device memory is released but not all backends consistently release the actual ggml_backend_buffer_t in system RAM, causing a memory leak. Bug #10040
[ { "path": "llama/patches/0001-cuda.patch", "patch": "@@ -1,47 +0,0 @@\n-From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001\n-From: jmorganca <jmorganca@gmail.com>\n-Date: Thu, 6 Jun 2024 23:55:47 -0700\n-Subject: [PATCH] cuda\n-\n----\n- ggml/src/ggml-backend.cpp | 1 -\n- ggml/src...
2025-04-14T19:12:36
huggingface/transformers
b33235e3c03a96ef5fc2cc084e368b658062ffa7
760eae30588036f09614ce09ec5d687faf1b595b
fix the error of tests/quantization/fbgemm_fp8/test_fbgemm_fp8.py::Fb… (#43547) fix the error of tests/quantization/fbgemm_fp8/test_fbgemm_fp8.py::FbgemmFp8Test::test_change_loading_attributes Signed-off-by: Wang, Yi <yi.a.wang@intel.com> Co-authored-by: Marc Sun <57196510+SunMarc@users.noreply.github.com> Co-authore...
[ { "path": "src/transformers/integrations/fbgemm_fp8.py", "patch": "@@ -310,12 +310,6 @@ def replace_with_fbgemm_fp8_linear(\n if new_module is None:\n continue\n \n- if hasattr(new_module, \"input_scale_ub\"):\n- new_module.input_scale_ub = torch.tensor(\n- ...
2026-02-10T14:36:35