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
denoland/deno
1ea5ff00c943fbca3c0b850118460fb21d4776e6
4650330ddb81f5a7622540e4560d9f944dac9aca
fix(ext/node): emit "connect" event on http.Server for CONNECT requests (#32599) ## Summary - Deno's `node:http` Server polyfill never emitted the `"connect"` event for HTTP CONNECT method requests - This broke HTTP proxy libraries (e.g. `proxy-chain`, used by Crawlee/Playwright) that rely on this event to handle tun...
[ { "path": "ext/http/00_serve.ts", "patch": "@@ -29,6 +29,8 @@ import {\n op_http_set_response_trailers,\n op_http_try_wait,\n op_http_upgrade_raw,\n+ op_http_upgrade_raw_connect,\n+ op_http_upgrade_raw_get_head,\n op_http_upgrade_websocket_next,\n op_http_wait,\n } from \"ext:core/ops\";\n@@ -15...
2026-03-12T21:33:40
ggml-org/llama.cpp
65e96a246490b4095acc11869da1d5300f89a493
9496bbb8081463b16412defcb4ef1ed17f7fb42f
docs : fix links in parsing.md (#18245) This commit corrects the links in the parsing.md which currently result in 404 errors.
[ { "path": "docs/development/parsing.md", "patch": "@@ -55,7 +55,7 @@ auto parser = build_chat_peg_native_parser([&](common_chat_peg_native_builder &\n ```\n \n For a more complete example, see `test_example_native()` in\n-[tests/test-chat-peg-parser.cpp](tests/test-chat-peg-parser.cpp).\n+[tests/test-chat-p...
2025-12-21T08:35:40
ollama/ollama
f4a73d57a4e3182a4dd25dae93a779a13423dfc0
948114e3e3325af01c31697dee861459624ca805
fix runner expire during active use. Clearing the expire timer as it is used. Allowing the finish to assign an expire timer so that the runner will expire after no use.
[ { "path": "server/sched.go", "patch": "@@ -296,6 +296,10 @@ func (pending *LlmRequest) useLoadedRunner(runner *runnerRef, finished chan *Llm\n \trunner.refMu.Lock()\n \tdefer runner.refMu.Unlock()\n \trunner.refCount++\n+\tif runner.expireTimer != nil {\n+\t\trunner.expireTimer.Stop()\n+\t\trunner.expireTim...
2024-04-28T16:43:40
vuejs/vue
dc97a39c2f41ce57431d42d8b41811866f8e105c
013d98092868a0c6721831e91616c64f99119b74
fix(codegen): support filters with () in older browsers (#7545) Fix #7544 Make sure no extra , is added at the end of the call so it also work with older browsers
[ { "path": "src/compiler/parser/filter-parser.js", "patch": "@@ -92,6 +92,6 @@ function wrapFilter (exp: string, filter: string): string {\n } else {\n const name = filter.slice(0, i)\n const args = filter.slice(i + 1)\n- return `_f(\"${name}\")(${exp},${args}`\n+ return `_f(\"${name}\")(${ex...
2018-03-08T16:24:09
denoland/deno
3d05945878003ef2db2763c97298a913d05645b2
8cf29d5da0c90654922ca35769eb6b11dbdb5a50
fix(ext/node): use non-blocking write in fs.writev (#32673) fs.writev inner async function was calling io.writeSync (blocking) instead of io.write (async). \ This switches to the non-blocking variant. Also, enables 4 node compat tests: - test-fs-writev.js - test-fs-writev-promises.js - test-fs-writev-sync.js - test-p...
[ { "path": "ext/node/polyfills/fs.ts", "patch": "@@ -2096,7 +2096,7 @@ function writev(\n let currentOffset = 0;\n // deno-lint-ignore prefer-primordials\n while (currentOffset < buffer.byteLength) {\n- currentOffset += await io.writeSync(fd, buffer.subarray(currentOffset));\n+ currentO...
2026-03-12T21:01:15
huggingface/transformers
280c5d6d4e67ff05f6140f788379abe484a84e34
e920f94b4e588a75a9154fba7de2f59fb2aa0db3
[loading] Allow loading to happen without threading (#42619) * add env var * oupsi * fix * style * add test * style
[ { "path": "src/transformers/core_model_loading.py", "patch": "@@ -31,7 +31,7 @@\n \n from .integrations.accelerate import offload_weight\n from .integrations.tensor_parallel import ALL_PARALLEL_STYLES\n-from .utils import is_torch_greater_or_equal, logging\n+from .utils import is_env_variable_true, is_torch...
2025-12-04T16:45:29
vuejs/vue
013d98092868a0c6721831e91616c64f99119b74
733c1be7f5983cdd9e8089a8088b235ba21a4dee
fix(show): prevent transitions from starting on change truthy values (#7524) Closes #7523
[ { "path": "src/platforms/web/runtime/directives/show.js", "patch": "@@ -27,7 +27,7 @@ export default {\n \n update (el: any, { value, oldValue }: VNodeDirective, vnode: VNodeWithData) {\n /* istanbul ignore if */\n- if (value === oldValue) return\n+ if (!value === !oldValue) return\n vnode =...
2018-03-08T16:19:08
ollama/ollama
948114e3e3325af01c31697dee861459624ca805
a3e60d9058d4459aac3a06f7cab80b0e894e1b38
fix sched to wait for the runner to terminate to ensure following vram check will be more accurate
[ { "path": "llm/server.go", "patch": "@@ -899,7 +899,10 @@ func (s *llmServer) Detokenize(ctx context.Context, tokens []int) (string, error\n func (s *llmServer) Close() error {\n \tif s.cmd != nil {\n \t\tslog.Debug(\"stopping llama server\")\n-\t\treturn s.cmd.Process.Kill()\n+\t\tif err := s.cmd.Process.K...
2024-04-28T16:41:38
ggml-org/llama.cpp
ddcb75dd8ac42dc23eb84f13bb17670fe9f2d49b
52ab19df633f3de5d4db171a16f2d9edd2342fec
server: add auto-sleep after N seconds of idle (#18228) * implement sleeping at queue level * implement server-context suspend * add test * add docs * optimization: add fast path * make sure to free llama_init * nits * fix use-after-free * allow /models to be accessed during sleeping, fix use-after-free * don...
[ { "path": "common/arg.cpp", "patch": "@@ -2887,6 +2887,16 @@ common_params_context common_params_parser_init(common_params & params, llama_ex\n params.lora_init_without_apply = true;\n }\n ).set_examples({LLAMA_EXAMPLE_SERVER}));\n+ add_opt(common_arg(\n+ {\"--sleep-idle-se...
2025-12-21T01:24:42
denoland/deno
2353a5abeb584ad0a8a2fc5a68ed80a9ab058628
e8a3cfafb61ea6cb31b65df63761c73b8dee0f98
fix(ext/node): wrap non-Error unhandled rejections in ERR_UNHANDLED_REJECTION (#32535) ## Summary - When a promise is rejected with a non-Error value (e.g. `Promise.reject(null)`) and there are no `unhandledRejection` listeners, Node.js wraps the rejection reason in an `ERR_UNHANDLED_REJECTION` error before routing i...
[ { "path": "ext/node/polyfills/process.ts", "patch": "@@ -1086,7 +1086,29 @@ function synchronizeListeners() {\n // listeners.\n \n event.preventDefault();\n- uncaughtExceptionHandler(event.reason, \"unhandledRejection\");\n+\n+ let reason = event.reason;\n+ // If the rej...
2026-03-12T18:08:51
huggingface/transformers
fda2d7350d717a6feb820065dad91e4302066070
b0831697be1dab7b88cccd95231ba3fbc50d90ee
feat(trainer): Just-in-time (JIT) asynchronous checkpointing using SIGTERM signals (#41723) * Just-in-time (JIT) asynchronous checkpointing using SIGTERM signals and cuda streams. * Fix failing ci tests * Update JIT checkpoint code to remove CUDA streams and async checkpointing. Introduce sentinal file to identify i...
[ { "path": "src/transformers/trainer.py", "patch": "@@ -642,6 +642,16 @@ def __init__(\n \"You should subclass `Trainer` and override the `create_optimizer_and_scheduler` method.\"\n )\n default_callbacks = DEFAULT_CALLBACKS + get_reporting_integration_callbacks(self.args....
2025-12-04T15:36:29
ggml-org/llama.cpp
408616adbdae2494b8bf23e048ef059fb681a474
9e39a1e6a991331bfa02390784eaa1ea226f1d4b
server : [easy] fix per round speculative decode logging (#18211) Currently we always log 0, as we clear slot.drafted before. To reproduce: Run llama-server with devstral-2 as main model and devstral-2-small as md, and verbose logging: ``` % ./build/bin/llama-server -v \ -m ~/llms/Devstral-2-123B-Instruct-2512-UD...
[ { "path": "tools/server/server-context.cpp", "patch": "@@ -2628,7 +2628,7 @@ struct server_context_impl {\n }\n }\n \n- SLT_DBG(slot, \"accepted %d/%d draft tokens, new n_tokens = %d\\n\", (int) ids.size() - 1, (int) slot.drafted.size(), slot.prompt.n_token...
2025-12-20T09:57:40
ollama/ollama
a3e60d9058d4459aac3a06f7cab80b0e894e1b38
f0c454ab577f4cf716e23ff56cbc3c316c88a78c
README.md: fix typos (#4007) Co-authored-by: Blake Mizerany <blake.mizerany@gmail.com>
[ { "path": "README.md", "patch": "@@ -1,5 +1,5 @@\n <div align=\"center\">\n- <img alt=\"ollama\" height=\"200px\" src=\"https://github.com/ollama/ollama/assets/3325447/0d0b44e2-8f4a-4e99-9b52-a5c1c741c8f7\">\n+ <img alt=\"ollama\" height=\"200px\" src=\"https://github.com/ollama/ollama/assets/3325447/0d0b4...
2024-05-01T17:39:38
vuejs/vue
733c1be7f5983cdd9e8089a8088b235ba21a4dee
e98d12d232af49b8e127375b6237013e01b77dda
fix(inject): use hasOwn instead of 'in' for provideKey check (#7460) fix #7284 * fix(Injection with Symbol polyfill): hasOwn instead of 'in' Symbol polyfill adds a setter on the Object prototype so the 'in' check evaluated to true on every object * test(Injected properties): Ensures prototype properties aren...
[ { "path": "src/core/instance/inject.js", "patch": "@@ -3,6 +3,7 @@\n import { warn } from '../util/index'\n import { hasSymbol } from 'core/util/env'\n import { defineReactive, observerState } from '../observer/index'\n+import { hasOwn } from 'shared/util'\n \n export function initProvide (vm: Component) {\...
2018-03-08T16:15:35
denoland/deno
e8a3cfafb61ea6cb31b65df63761c73b8dee0f98
b31680f5515c3a6bdadce1d64f6bf61a1aaa4bd8
chore(npm): fix debug panic in DepEntryCache (#32668) Closes https://github.com/denoland/deno/issues/32622
[ { "path": "libs/npm/resolution/graph.rs", "patch": "@@ -1050,17 +1050,19 @@ struct DepEntryCache(HashMap<Rc<PackageNv>, Rc<Vec<NpmDependencyEntry>>>);\n impl DepEntryCache {\n pub fn store(\n &mut self,\n- nv: Rc<PackageNv>,\n+ nv: &Rc<PackageNv>,\n version_info: &NpmPackageVersionInfo,\n ...
2026-03-12T17:21:28
ggml-org/llama.cpp
9e39a1e6a991331bfa02390784eaa1ea226f1d4b
74e05131e98661d56b30a1a9a1165165a0c70230
server: support load model on startup, support preset-only options (#18206) * server: support autoload model, support preset-only options * add docs * load-on-startup * fix * Update common/arg.cpp Co-authored-by: Pascal <admin@serveurperso.com> --------- Co-authored-by: Pascal <admin@serveurperso.com>
[ { "path": "common/arg.cpp", "patch": "@@ -96,6 +96,11 @@ common_arg & common_arg::set_sparam() {\n return *this;\n }\n \n+common_arg & common_arg::set_preset_only() {\n+ is_preset_only = true;\n+ return *this;\n+}\n+\n bool common_arg::in_example(enum llama_example ex) {\n return examples.find...
2025-12-20T08:25:27
huggingface/transformers
ce53cc00aac9fac2669edf1ff221fb146d577a61
a3e2d54755658bc80ccb70f8578b58ca9c447a97
[V5] Return a BatchEncoding dict from apply_chat_template by default again (#42567) * Flip the default return type for `apply_chat_template` to match the underlying tokenizer * Remove test_tokenization_for_chat tests, which no longer do anything useful * Remove test_tokenization_for_chat tests, which no longer do an...
[ { "path": "src/transformers/tokenization_utils_base.py", "patch": "@@ -3195,7 +3195,7 @@ def apply_chat_template(\n truncation: bool = False,\n max_length: Optional[int] = None,\n return_tensors: Optional[Union[str, TensorType]] = None,\n- return_dict: bool = False,\n+ ...
2025-12-04T14:44:19
ollama/ollama
bd8eed57fc7d5f9b9b9d333b9c395864fb2378d8
9cf0f2e9736c31bbd00f4671613d0e31ecb3c4ea
fix parser name
[ { "path": "parser/parser.go", "patch": "@@ -27,8 +27,9 @@ const (\n )\n \n var (\n-\terrMissingFrom = errors.New(\"no FROM line\")\n-\terrInvalidRole = errors.New(\"role must be one of \\\"system\\\", \\\"user\\\", or \\\"assistant\\\"\")\n+\terrMissingFrom = errors.New(\"no FROM line\")\n+\terrInval...
2024-04-27T00:11:47
vuejs/vue
80e650cfdf9f8127405592d64ba5af1e6e13ce86
9b22d86ab315a3c6061a6a4776eab1964304f92e
chore: fix formatting
[ { "path": "test/unit/modules/compiler/compiler-options.spec.js", "patch": "@@ -9,17 +9,18 @@ describe('compile options', () => {\n <input type=\"text\" v-model=\"msg\" required max=\"8\" v-validate:field1.group1.group2>\n </div>\n `, {\n- directives: {\n- validate (el, dir)...
2018-03-08T14:52:27
denoland/deno
b31680f5515c3a6bdadce1d64f6bf61a1aaa4bd8
7698bccccc54241d8be1c50cd22b3c4af0360998
fix(runtime): reduce memory retention after web worker termination (#32617) ## Summary Addresses #26058 — Web Workers use significantly more RSS than Chrome, and terminating them doesn't release the memory back to the OS. Two targeted changes: - **Call `malloc_trim(0)` on Linux after each worker thread exits.** Whe...
[ { "path": "runtime/ops/worker_host.rs", "patch": "@@ -269,7 +269,20 @@ fn op_create_worker(\n .await\n };\n \n- create_and_run_current_thread(fut)\n+ let _ = create_and_run_current_thread(fut);\n+\n+ // After the worker's tokio runtime and JsRuntime/V8 isolate have been\n+ // dropped, ...
2026-03-12T17:01:09
ggml-org/llama.cpp
14931a826e4f5b4536f03f65c8d568d99bf64f0e
f99ef53d2a9591a9a32d28f402630cf7c58f1298
arg: fix order to use short form before long form (#18196) * arg: fix order to use short form before long form * arg: update doc * arg: update test-arg-parser * arg: address review feedback from ngxson simplified to check first.length() <= last.length() only fixed: --sampler-seq, --rerank, --draft ordering note: m...
[ { "path": "common/arg.cpp", "patch": "@@ -1144,15 +1144,15 @@ common_params_context common_params_parser_init(common_params & params, llama_ex\n }\n ).set_env(\"LLAMA_ARG_CTX_CHECKPOINTS\").set_examples({LLAMA_EXAMPLE_SERVER, LLAMA_EXAMPLE_CLI}));\n add_opt(common_arg(\n- {\"--cache-r...
2025-12-19T17:01:56
huggingface/transformers
a3e2d54755658bc80ccb70f8578b58ca9c447a97
bf8b9e7b53ec723257f07ef5602b77db2fe49bff
Fix typo in docstring in modeling_sam3_tracker.py (#42438) * Update modeling_sam3_tracker.py * Apply docstring to modular_sam3_tracker.py for consistency Added custom introduction to auto_docstring for Sam3TrackerPreTrainedModel. to match typo fix in modeling_sam3_tracker.py * Fix SAM3 tracker docstrings in modula...
[ { "path": "src/transformers/models/sam3_tracker/modeling_sam3_tracker.py", "patch": "@@ -107,7 +107,12 @@ def forward(self, hidden_states):\n return hidden_states\n \n \n-@auto_docstring\n+@auto_docstring(\n+ custom_intro=\"\"\"\n+ Segment Anything Model 3 (SAM 3) for generating segmentation m...
2025-12-04T14:36:27
ollama/ollama
8907bf51d235e86854c762870f203380866f1ae3
abe614c705736eed06440d657ab75ca094fc78f3
fix multiline
[ { "path": "parser/parser.go", "patch": "@@ -110,11 +110,16 @@ func Parse(r io.Reader) (cmds []Command, err error) {\n \tcase stateComment, stateNil:\n \t\t// pass; nothing to flush\n \tcase stateValue:\n-\t\tif _, ok := unquote(b.String()); !ok {\n+\t\ts, ok := unquote(b.String())\n+\t\tif !ok {\n \t\t\tret...
2024-04-25T02:17:26
ggml-org/llama.cpp
cc0a04343e56f2decdf2a6df977ff9c9edc85ced
98c1c7a7bf433f1f64dee23f25e7237caab868c9
server: friendlier error msg when ctx < input (#18174) * llama-server: friendlier error msg when ctx < input This PR adds formatted strings to the server's send_error function * llama-server: use string_format inline * fix test
[ { "path": "tools/server/server-context.cpp", "patch": "@@ -1974,19 +1974,33 @@ struct server_context_impl {\n \n if (!slot.can_split()) {\n if (slot.task->n_tokens() > n_ubatch) {\n- send_error(slot, \"input is too large to p...
2025-12-19T11:10:00
vuejs/vue
9b22d86ab315a3c6061a6a4776eab1964304f92e
3d431a5ddc62fac77a0d0466e13427dd7985901d
fix: install ssr helpers for functional context during SSR close #7443, ref nuxt/nuxt.js#2565
[ { "path": "flow/component.js", "patch": "@@ -16,6 +16,8 @@ declare interface Component {\n static directive: (id: string, def?: Function | Object) => Function | Object | void;\n static component: (id: string, def?: Class<Component> | Object) => Class<Component>;\n static filter: (id: string, def?: Fun...
2018-03-08T14:51:03
denoland/deno
7698bccccc54241d8be1c50cd22b3c4af0360998
9eed28a85df7d142be3ab93b5072d9dcdce2acd0
fix(runtime): display proper error when throwing in event handlers (#32663) ## Summary - Fixes `"Uncaught null"` being displayed instead of the actual error when throwing in `unload`, `beforeunload`, `load`, or process `exit`/`beforeExit` event handlers - Root cause: `reportException` in JS calls `op_dispatch_excepti...
[ { "path": "libs/core/error.rs", "patch": "@@ -1313,7 +1313,7 @@ pub(crate) fn exception_to_err_result<'s, 'i, T>(\n Err(exception_to_err(scope, exception, in_promise, clear_error))\n }\n \n-pub(crate) fn exception_to_err<'s, 'i>(\n+pub fn exception_to_err<'s, 'i>(\n scope: &mut v8::PinScope<'s, 'i>,\n ...
2026-03-12T16:42:00
huggingface/transformers
bf8b9e7b53ec723257f07ef5602b77db2fe49bff
552409e598e91857923b44143934b02e78e82f57
Fix _is_package_available to handle underscore/hyphen equivalence (#42615) Fix minibug
[ { "path": "src/transformers/utils/import_utils.py", "patch": "@@ -55,9 +55,15 @@ def _is_package_available(pkg_name: str, return_version: bool = False) -> tuple[\n # importlib.metadata works with the distribution package, which may be different from the import\n # name (e.g. `PIL` is...
2025-12-04T13:52:33
ollama/ollama
0a7fdbe533b087c47079a2cc54de6bf916443cbc
5950c176caf585b04ac517003d1385833e5e4927
prompt to display and add local ollama keys to account (#3717) - return descriptive error messages when unauthorized to create blob or push a model - display the local public key associated with the request that was denied
[ { "path": "auth/auth.go", "patch": "@@ -10,12 +10,44 @@ import (\n \t\"log/slog\"\n \t\"os\"\n \t\"path/filepath\"\n+\t\"strings\"\n \n \t\"golang.org/x/crypto/ssh\"\n )\n \n const defaultPrivateKey = \"id_ed25519\"\n \n+func keyPath() (string, error) {\n+\thome, err := os.UserHomeDir()\n+\tif err != nil {\...
2024-04-30T18:02:08
vuejs/vue
1c0b4af5fd2f9e8173b8f4718018ee80a6313872
f7ca21eab1e0d661945aa6070fc988028c90966f
fix: skip v-model & value binding collision check with dynamic type binding (#7406) * fix #7404
[ { "path": "src/platforms/web/compiler/directives/model.js", "patch": "@@ -130,9 +130,11 @@ function genDefaultModel (\n const type = el.attrsMap.type\n \n // warn if v-bind:value conflicts with v-model\n+ // except for inputs with v-bind:type\n if (process.env.NODE_ENV !== 'production') {\n const...
2018-03-07T22:44:55
ggml-org/llama.cpp
98c1c7a7bf433f1f64dee23f25e7237caab868c9
acb73d83409eb1511803f9df94ac3aa3eb36c50a
presets: refactor, allow cascade presets from different sources, add global section (#18169) * presets: refactor, allow cascade presets from different sources * update docs * fix neg arg handling * fix empty mmproj * also filter out server-controlled args before to_ini() * skip loading custom_models if not specif...
[ { "path": "common/arg.cpp", "patch": "@@ -772,6 +772,11 @@ bool common_params_to_map(int argc, char ** argv, llama_example ex, std::map<com\n }\n auto opt = *arg_to_options[arg];\n std::string val;\n+ if (opt.value_hint == nullptr && opt.value_hint_2 == nullptr) {\n+ ...
2025-12-19T11:08:20
denoland/deno
aa6ee7d5bbf4ae534e101ccdd86102284d4f9fcd
5b9986dd605081f6a8e04b4f8d2fcdba26bc19ec
fix(ext/napi): call wrap/ref finalizers at shutdown (#32592) - Track NAPI finalizer callbacks (from `napi_wrap`, `napi_create_external`, `napi_add_finalizer`, `napi_set_instance_data`) in a `PendingNapiFinalizer` list shared between `NapiState` and `Env` - Call tracked finalizers in LIFO order during `MainWorker` shut...
[ { "path": "cli/lib/worker.rs", "patch": "@@ -853,6 +853,11 @@ impl LibMainWorker {\n self.worker.dispatch_process_exit_event()\n }\n \n+ #[inline]\n+ pub fn run_napi_ref_finalizers(&mut self) {\n+ self.worker.run_napi_ref_finalizers()\n+ }\n+\n pub async fn execute_main_module(&mut self) -> Re...
2026-03-12T16:07:58
huggingface/transformers
552409e598e91857923b44143934b02e78e82f57
bebfab0682839718443c7bf5020567ab3b434057
Fix small weight loading example (#42622) fix
[ { "path": "src/transformers/core_model_loading.py", "patch": "@@ -694,7 +694,7 @@ def convert_and_load_state_dict_in_model(\n target_patterns=[\"q\", \"k\",\"v\"],\n operations=[Chunk(dim=0, chunks=3)]),\n collected_tensors={\n- \"qkv\": [Fu...
2025-12-04T13:27:01
vuejs/vue
f7ca21eab1e0d661945aa6070fc988028c90966f
3eb37acf98e2d9737de897ebe7bdb7e9576bcc21
fix: always install composition event listeners Previously the installation was skipped on Android because it was not needed for Chinese IME - however some IMEs such as Japanese exhibits the same behavior as on other browers. So it is safer to always enable the check. Closes #7367
[ { "path": "src/platforms/web/runtime/directives/model.js", "patch": "@@ -6,7 +6,7 @@\n import { isTextInputType } from 'web/util/element'\n import { looseEqual, looseIndexOf } from 'shared/util'\n import { mergeVNodeHook } from 'core/vdom/helpers/index'\n-import { warn, isAndroid, isIE9, isIE, isEdge } from...
2018-03-07T21:50:38
ggml-org/llama.cpp
0a271d82b46577d955b6c5c8020aab3b4d21057d
52fc7fee8a96211de439aa8ea27dd53e7a4a2200
model-conversion : add verbose flag in run-org-model.py (#18194) This commit adds a --verbose flag to the run-org-model.py script to enable or disable detailed debug output, such as input and output tensors for each layer. Debug utilities (summarize, debug_hook, setup_rope_debug) have been moved to utils/common.py. T...
[ { "path": "examples/model-conversion/scripts/causal/run-org-model.py", "patch": "@@ -2,135 +2,22 @@\n \n import argparse\n import os\n+import sys\n import importlib\n from pathlib import Path\n \n+# Add parent directory to path for imports\n+sys.path.insert(0, os.path.join(os.path.dirname(__file__), '..'))\...
2025-12-19T07:43:16
denoland/deno
5b9986dd605081f6a8e04b4f8d2fcdba26bc19ec
c0c896b17a62f37a81e30865069307db7d629720
fix(ext/node): improve crypto.generateKeyPair validation (#32620) ## Summary - Restructures `generateKeyPair` to call `createJob` synchronously so validation errors throw synchronously (matching Node.js behavior) - Adds key encoding validation (format, type, cipher, passphrase) with proper Node.js error codes (`ERR_IN...
[ { "path": "ext/node/polyfills/internal/crypto/cipher.ts", "patch": "@@ -594,10 +594,30 @@ function _lazyInitDecipherDecoder(self: any, encoding: string) {\n }\n }\n \n+const ENCRYPT_UNSUPPORTED_KEY_TYPES = new Set([\n+ \"rsa-pss\",\n+ \"dsa\",\n+ \"ec\",\n+ \"ed25519\",\n+ \"ed448\",\n+ \"x25519\",\...
2026-03-12T16:05:27
huggingface/transformers
bebfab0682839718443c7bf5020567ab3b434057
0c3d043e659e022146a84c8931a5f2b06543f29b
[core] Fix quark (#42457) * intial * update * add convert * fix * style * rm comment * explain * loop * fix * fix * update * Apply style fixes * fix * style * update --------- Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
[ { "path": "src/transformers/core_model_loading.py", "patch": "@@ -462,9 +462,15 @@ def convert(\n full_name = layer_name\n if \".*.\" in layer_name:\n full_name = layer_name.replace(\".*.\", \".0.\")\n- prefix, _, suffix = next(full_name.partition(k) for k in collected_ten...
2025-12-04T12:30:41
ollama/ollama
7b59d1770fdcece7ee660dc0c0f1752c7b773e2f
95ead8ffba312524c233063f89635953c44f2a73
Fix relative path lookup
[ { "path": "gpu/assets.go", "patch": "@@ -40,7 +40,7 @@ func PayloadsDir() (string, error) {\n \t\t\t}\n \n \t\t\tvar paths []string\n-\t\t\tfor _, root := range []string{appExe, cwd} {\n+\t\t\tfor _, root := range []string{filepath.Dir(appExe), cwd} {\n \t\t\t\tpaths = append(paths,\n \t\t\t\t\tfilepath.Joi...
2024-04-29T23:00:08
vuejs/vue
3eb37acf98e2d9737de897ebe7bdb7e9576bcc21
aeebf0de906e7030282d59734ad30d69548cb521
fix: fix ssr env detection in weex (#7375)
[ { "path": "src/core/util/env.js", "patch": "@@ -38,7 +38,7 @@ let _isServer\n export const isServerRendering = () => {\n if (_isServer === undefined) {\n /* istanbul ignore if */\n- if (!inBrowser && typeof global !== 'undefined') {\n+ if (!inBrowser && !inWeex && typeof global !== 'undefined') ...
2018-03-07T20:37:14
ggml-org/llama.cpp
52fc7fee8a96211de439aa8ea27dd53e7a4a2200
cdbada8d102c32bead62e9c26473cc1ea37939f0
android: fix missing screenshots for Android.md (#18156) * Android basic sample app layout polish * Add missing screenshots and polish android README doc * Replace file blobs with URLs served by GitHub pages service.
[ { "path": "docs/android.md", "patch": "@@ -1,27 +1,27 @@\n \n # Android\n \n-## Build with Android Studio\n+## Build GUI binding using Android Studio\n \n Import the `examples/llama.android` directory into Android Studio, then perform a Gradle sync and build the project.\n-![Project imported into Android St...
2025-12-19T07:32:04
huggingface/transformers
0c3d043e659e022146a84c8931a5f2b06543f29b
9b74e4c43402049608fc3e74c9b675d9b37ef0f7
fix(Qwen3VLCausalLMOutputWithPast): missing `hidden_states` and `atte… (#42609) * fix(Qwen3VLCausalLMOutputWithPast): missing `hidden_states` and `attentions` kwargs * revert kwargs for the base class * make regenerated model files * symmetrical change for `modular_qwen3_vl_moe.py` * regenerated `modeling_qwen3_vl...
[ { "path": "src/transformers/models/qwen3_vl/modeling_qwen3_vl.py", "patch": "@@ -38,7 +38,7 @@\n from ...modeling_rope_utils import ROPE_INIT_FUNCTIONS, dynamic_rope_update\n from ...modeling_utils import ALL_ATTENTION_FUNCTIONS, PreTrainedModel\n from ...processing_utils import Unpack\n-from ...utils impor...
2025-12-04T11:53:39
vuejs/vue
7392dfcc1d5fd7b257df5ae134f9eb2f0cc0a51e
f7311c90b35da6ee680dacb0566a276bff907ccb
fix(observer): do not invoke getters on initial observation (#7302) fix #7280
[ { "path": "src/core/observer/index.js", "patch": "@@ -61,7 +61,7 @@ export class Observer {\n walk (obj: Object) {\n const keys = Object.keys(obj)\n for (let i = 0; i < keys.length; i++) {\n- defineReactive(obj, keys[i], obj[keys[i]])\n+ defineReactive(obj, keys[i])\n }\n }\n \n@@ ...
2018-03-07T19:58:12
ollama/ollama
95ead8ffba312524c233063f89635953c44f2a73
7aa08a77caf0121b9c9be8df1922b5fb8679d377
Restart server on failure when running Windows app (#3985) * app: restart server on failure * fix linter * address comments * refactor log directory creation to be where logs are written * check all log dir creation errors
[ { "path": "app/lifecycle/server.go", "patch": "@@ -43,37 +43,36 @@ func getCLIFullPath(command string) string {\n \treturn command\n }\n \n-func SpawnServer(ctx context.Context, command string) (chan int, error) {\n-\tdone := make(chan int)\n-\n-\tlogDir := filepath.Dir(ServerLogFile)\n-\t_, err := os.Stat(...
2024-04-29T14:07:52
denoland/deno
c0c896b17a62f37a81e30865069307db7d629720
089589763706db9b1915b605918b2bd81f487e53
fix(cli): update typescript-go to 2026-03-09 (#32656)
[ { "path": "cli/build.rs", "patch": "@@ -26,7 +26,6 @@ fn compress_decls(out_dir: &Path) {\n \"lib.deno_cache.d.ts\",\n \"lib.deno_net.d.ts\",\n \"lib.deno_broadcast_channel.d.ts\",\n- \"lib.temporal.d.ts\",\n \"lib.decorators.d.ts\",\n \"lib.decorators.legacy.d.ts\",\n \"lib.dom.a...
2026-03-12T13:55:02
ggml-org/llama.cpp
8ea958d4d9d5bd43f9f244950bf057b5227ce8d9
f9ec8858edea4a0ecfea149d6815ebfb5ecc3bcd
model : add ASR support for LFM2-Audio-1.5B (conformer) (#18106) * ASR with LFM2-Audio-1.5B * Set rope_theta * Fix comment * Remove rope_theta setting * Address PR feedback * rename functions to conformer * remove some redundant ggml_cont * fix missing tensor * add prefix "a." for conv tensors * remove redund...
[ { "path": "common/arg.cpp", "patch": "@@ -1196,7 +1196,7 @@ common_params_context common_params_parser_init(common_params & params, llama_ex\n [](common_params & params, const std::string & value) {\n params.system_prompt = value;\n }\n- ).set_examples({LLAMA_EXAMPLE_COMPLETIO...
2025-12-18T23:18:01
vuejs/vue
1c8e2e88ed2d74a02178217b318564b73a096c18
50b711af43708426e63b4ea529436b49fafc3f2e
fix: support KeyboardEvent.key in built-in keyboard event modifiers (#7121) fix #6900
[ { "path": "src/core/instance/render-helpers/check-keycodes.js", "patch": "@@ -3,6 +3,26 @@\n import config from 'core/config'\n import { hyphenate } from 'shared/util'\n \n+const keyNames: { [key: string]: string | Array<string> } = {\n+ esc: 'Escape',\n+ tab: 'Tab',\n+ enter: 'Enter',\n+ space: ' ',\n+...
2018-03-07T19:38:41
ollama/ollama
586672f4906c310d6321d42176b32bf9430fa031
b03408de741a9e6b22025216a15657b2cb4fbfe5
fix copying model to itself (#4019)
[ { "path": "server/images.go", "patch": "@@ -710,6 +710,10 @@ func CopyModel(src, dst model.Name) error {\n \t\treturn model.Unqualified(src)\n \t}\n \n+\tif src.Filepath() == dst.Filepath() {\n+\t\treturn nil\n+\t}\n+\n \tmanifests, err := GetManifestPath()\n \tif err != nil {\n \t\treturn err", "additi...
2024-04-29T03:47:49
denoland/deno
089589763706db9b1915b605918b2bd81f487e53
bbd5282f849f0b1372d55594fc0083d668f3a4d3
fix: wake event loop when V8 posts foreground tasks from background threads (#32450) When V8 background threads post foreground tasks (module compilation callbacks, Atomics.waitAsync timeouts, worker lifecycle events), the event loop had no way to know it should wake up and pump those tasks. This caused stalls where p...
[ { "path": "Cargo.lock", "patch": "@@ -11161,9 +11161,9 @@ dependencies = [\n \n [[package]]\n name = \"v8\"\n-version = \"146.3.0\"\n+version = \"146.4.0\"\n source = \"registry+https://github.com/rust-lang/crates.io-index\"\n-checksum = \"8741c1f33d1ebf6c81f53fbb0c66fdd7a2b55feeead0ec40ec9468b387d67176\"\n...
2026-03-12T13:36:08
ggml-org/llama.cpp
f716588e63224b2f33bb5d13b397fbcfabefa888
4d1316c440e7ee7b854423aea8db6ffcd92caeaf
ggml-cpu: extend support for RVV floating-point kernels (#17318) * cmake: add BF16 RVV flag for ggml-cpu * ggml-cpu: add floating-point conversion kernels * ggml: add floating-point kernels Co-authored-by: Rehan Qasim <rehan.qasim@10xengineers.ai> * ggml-cpu: fix lmul in vec_dot_bf16 * ggml-cpu: change redsum to ...
[ { "path": "ggml/src/ggml-cpu/CMakeLists.txt", "patch": "@@ -458,13 +458,17 @@ function(ggml_add_cpu_backend_variant_impl tag_name)\n if (GGML_RV_ZFH)\n string(APPEND MARCH_STR \"_zfh\")\n endif()\n+\n if (GGML_XTHEADVECTOR)\n string(APPEND ...
2025-12-18T14:02:09
huggingface/transformers
9b74e4c43402049608fc3e74c9b675d9b37ef0f7
3cdccba080c130805dcfabda4e9baa49802853b7
Every model forward() should have **kwargs (#42603) * Add **kwargs into every Model.forward() * Add the test back in * And the others I missed * Fix udop test * Fix fast2speech2conformer test * make fixup
[ { "path": "src/transformers/models/align/modeling_align.py", "patch": "@@ -1004,6 +1004,7 @@ def forward(\n pixel_values: Optional[torch.FloatTensor] = None,\n output_hidden_states: Optional[bool] = None,\n return_dict: Optional[bool] = None,\n+ **kwargs,\n ) -> Union[tupl...
2025-12-04T10:51:45
vuejs/vue
50b711af43708426e63b4ea529436b49fafc3f2e
c57ffb768def2b8a409472eab2715579fc3f047d
fix: do not special case attributes for custom elements close #6864, close #6885
[ { "path": "src/platforms/web/runtime/modules/attrs.js", "patch": "@@ -59,7 +59,9 @@ function updateAttrs (oldVnode: VNodeWithData, vnode: VNodeWithData) {\n }\n \n function setAttr (el: Element, key: string, value: any) {\n- if (isBooleanAttr(key)) {\n+ if (el.tagName.indexOf('-') > -1) {\n+ baseSetAtt...
2018-03-07T17:03:31
ggml-org/llama.cpp
4d1316c440e7ee7b854423aea8db6ffcd92caeaf
ec7b9329aed15302cf22528a286df673f09e4280
arg: fix ASAN error on sampler_type_names empty (#18167)
[ { "path": "common/arg.cpp", "patch": "@@ -873,7 +873,9 @@ common_params_context common_params_parser_init(common_params & params, llama_ex\n sampler_type_chars += common_sampler_type_to_chr(sampler);\n sampler_type_names += common_sampler_type_to_str(sampler) + \";\";\n }\n- sampler_t...
2025-12-18T13:30:32
denoland/deno
bbd5282f849f0b1372d55594fc0083d668f3a4d3
f0dfb89720bf05d401d082bad87babe78ac036aa
fix(npm): support npmrc certfile + keyfile (#32655) This PR implements the missing support for `certfile` and `keyfile` options. Now when both are set, it creates an HTTP client with the client certificate attached.
[ { "path": "cli/http_util.rs", "patch": "@@ -1,8 +1,11 @@\n // Copyright 2018-2026 the Deno authors. MIT license.\n \n+use std::io::BufReader;\n+use std::path::Path;\n use std::sync::Arc;\n \n use boxed_error::Boxed;\n+use dashmap::DashMap;\n use deno_cache_dir::file_fetcher::RedirectHeaderParseError;\n use ...
2026-03-12T12:54:58
huggingface/transformers
2e93004ae0321af3188e2772676429a01436f9fc
a48d68c6b06172938948eef457ebb7bcd5dcb77c
extend FA2 and other cases to XPU, (#42536) * extend FA2 and other cases to XPU, we expect all model cases except CUDAGraph specific, CUDA compute capability specific and FA3 specific can run XPU. For FA3, we are develioping Signed-off-by: Yao, Matrix <matrix.yao@intel.com> * fix style Signed-off-by: Yao, Matrix <...
[ { "path": "tests/generation/test_continuous_batching.py", "patch": "@@ -25,7 +25,6 @@\n require_kernels,\n require_read_token,\n require_torch_accelerator,\n- require_torch_gpu,\n slow,\n torch_device,\n )\n@@ -315,36 +314,47 @@ def test_continuous_batching_parity_gemma_sdpa(self) -> ...
2025-12-04T07:39:20
vuejs/vue
c57ffb768def2b8a409472eab2715579fc3f047d
d6200d70261c4a8943190900e0721ede1c4a4f2b
chore: fix README.md vulnerable target="_blank" links (#7689) * README: add rel noopener and noreferrer Add to all "_blank" links rel noopener and noreferrer to prevent vulnerability * Update README.md
[ { "path": "README.md", "patch": "@@ -1,4 +1,4 @@\n-<p align=\"center\"><a href=\"https://vuejs.org\" target=\"_blank\"><img width=\"100\" src=\"https://vuejs.org/images/logo.png\" alt=\"Vue logo\"></a></p>\n+<p align=\"center\"><a href=\"https://vuejs.org\" target=\"_blank\" rel=\"noopener noreferrer\"><img...
2018-03-07T14:03:07
ollama/ollama
d6e3b64582d93dfa0c836965b70d66c8487c2f8b
114c932a8e872846fc714353c65d041feb886027
Fix concurrency for CPU mode Prior refactoring passes accidentally removed the logic to bypass VRAM checks for CPU loads. This adds that back, along with test coverage. This also fixes loaded map access in the unit test to be behind the mutex which was likely the cause of various flakes in the tests.
[ { "path": "server/sched.go", "patch": "@@ -149,6 +149,14 @@ func (s *Scheduler) processPending(ctx context.Context) {\n \t\t\t\t\t\tbreak\n \t\t\t\t\t}\n \n+\t\t\t\t\t// If we're CPU only mode, just limit by loadedMax above\n+\t\t\t\t\t// TODO handle system memory exhaustion\n+\t\t\t\t\tif (len(gpus) == 1 &...
2024-04-28T20:40:31
ggml-org/llama.cpp
9ce64aed7d74f6166dc17ad1d4bd544e76735418
900316da4e1443723d0d49835607c9fd218870e7
webui: Fix selecting generated output issues during active streaming (#18091) * draft: incremental markdown rendering with stable blocks * refactor: Logic improvements * refactor: DRY Markdown post-processing logic * refactor: ID generation improvements * fix: Remove runes * refactor: Clean up & add JSDocs * cho...
[ { "path": "tools/server/webui/package-lock.json", "patch": "@@ -2109,9 +2109,9 @@\n \t\t\t}\n \t\t},\n \t\t\"node_modules/@sveltejs/kit\": {\n-\t\t\t\"version\": \"2.48.5\",\n-\t\t\t\"resolved\": \"https://registry.npmjs.org/@sveltejs/kit/-/kit-2.48.5.tgz\",\n-\t\t\t\"integrity\": \"sha512-/rnwfSWS3qwUSzvHy...
2025-12-18T10:13:52
denoland/deno
9d246e77dd1afbdda4f21f7377ad6528048e4f7b
40b56f77c3798065de8150b4380549985ba83c51
fix(npm): retain authorization header on http->https redirect (#29878) Fixes #29869. ## Summary When an npm registry is configured with `http://` in `.npmrc` and the server redirects to `https://`, the authorization header was being stripped. This happened because the previous origin-based check (`url.origin() != ne...
[ { "path": "cli/http_util.rs", "patch": "@@ -351,8 +351,7 @@ impl HttpClient {\n let mut req = self.get(new_url.clone())?.build();\n \n let mut headers = headers.clone();\n- // SECURITY: Do NOT forward auth headers to a new origin\n- if new_url.origin() != url.origin() {\n+ ...
2026-03-12T11:16:14
huggingface/transformers
a48d68c6b06172938948eef457ebb7bcd5dcb77c
9e82c7791e0c94de7a536886456017b0eadcc08b
Fix some models cache initialization (#42586) * Create cache when training in case generate needs being called * Align modular * fixes * cohere * fix modular * fix * review --------- Co-authored-by: Cyril Vallez <cyril.vallez@gmail.com>
[ { "path": "src/transformers/models/cohere2/modeling_cohere2.py", "patch": "@@ -29,7 +29,6 @@\n from ...cache_utils import Cache, DynamicCache\n from ...generation import GenerationMixin\n from ...masking_utils import create_causal_mask, create_sliding_window_causal_mask\n-from ...modeling_flash_attention_ut...
2025-12-03T17:18:20
vuejs/vue
d6200d70261c4a8943190900e0721ede1c4a4f2b
990374bacbaa75ae2535370583714f1032e768a1
feat(weex): update weex recycle-list compiler (#7610) + Support v-once + Generate @templateId on the root element of each component + Add binding-expression attribute on recycle-list + Fix the compile result of v-else-if and v-else
[ { "path": "src/platforms/weex/compiler/modules/recycle-list/component-root.js", "patch": "@@ -10,6 +10,7 @@ export function postTransformComponentRoot (\n if (!el.parent) {\n // component root\n addAttr(el, '@isComponentRoot', 'true')\n+ addAttr(el, '@templateId', '_uid')\n addAttr(el, '@co...
2018-03-05T03:59:21
ggml-org/llama.cpp
900316da4e1443723d0d49835607c9fd218870e7
57c1e0564365b76a954304812cc90d3e60939bc7
webui: fix chat screen shadow width (#18010) * webui: fix chat screen shadow width * chore: add index.html.gz
[ { "path": "tools/server/webui/src/lib/components/app/chat/ChatScreen/ChatScreen.svelte", "patch": "@@ -587,7 +587,7 @@\n \n \t\t&::after {\n \t\t\tcontent: '';\n-\t\t\tposition: fixed;\n+\t\t\tposition: absolute;\n \t\t\tbottom: 0;\n \t\t\tz-index: -1;\n \t\t\tleft: 0;", "additions": 1, "deletions":...
2025-12-18T10:08:42
huggingface/transformers
9e82c7791e0c94de7a536886456017b0eadcc08b
d5d87934c2294da88802fdef7dbd22b89f38f2be
Fix Ernie Moe Test (#42595) * fix * fix * rm unnecessary config * remove references
[ { "path": "tests/models/ernie4_5_moe/test_modeling_ernie4_5_moe.py", "patch": "@@ -97,24 +97,27 @@ def test_load_balancing_loss(self):\n config, input_dict = self.model_tester.prepare_config_and_inputs_for_common()\n config.num_labels = 3\n config.num_experts = 3\n- config.exp...
2025-12-03T17:09:43
denoland/deno
40b56f77c3798065de8150b4380549985ba83c51
b8d9580c8eb528c581ef592ba6a014c9bfd29c19
perf(ext/web): reduce promise allocations in streams (#32652) ## Summary - Collapses `uponPromise`'s double `.then()` chain into a single `.then()` — the second chain only caught internal assertion errors, now handled via try/catch in wrapped handlers - Converts `setPromiseIsHandledToTrue(PromisePrototypeThen(...))` ...
[ { "path": "ext/web/06_streams.js", "patch": "@@ -220,10 +220,31 @@ function uponRejection(promise, onRejected) {\n * @returns {void}\n */\n function uponPromise(promise, onFulfilled, onRejected) {\n+ // Optimized: single .then() instead of double promise chain.\n+ // The original code was:\n+ // Prom...
2026-03-12T07:49:15
vuejs/vue
cd334070f3b82d3f5892c4999cc290ccd4f56fd8
1fa4a5e0fb2191e5b48b72710c92feaf51b3a979
fix: address potential regex backtrack
[ { "path": "src/compiler/codegen/events.js", "patch": "@@ -1,7 +1,7 @@\n /* @flow */\n \n-const fnExpRE = /^\\s*([\\w$_]+|\\([^)]*?\\))\\s*=>|^function\\s*\\(/\n-const simplePathRE = /^\\s*[A-Za-z_$][\\w$]*(?:\\.[A-Za-z_$][\\w$]*|\\['.*?']|\\[\".*?\"]|\\[\\d+]|\\[[A-Za-z_$][\\w$]*])*\\s*$/\n+const fnExpRE = ...
2018-02-21T18:11:11
ggml-org/llama.cpp
0a0bba05e8390ab7e4a54bb8c0ed0a25da64cf62
5166aaf86898c5d5268b2c4f770a35cb7ca1a0f6
ggml-hexagon: swiglu_oai operation (#18114) * snapshot: debug ggml-hexagon swiglu-oai * fix: fix hvx_min_scalar_f32 * feat: working swiglu-oai * chore: fix formating isue
[ { "path": "ggml/src/ggml-hexagon/ggml-hexagon.cpp", "patch": "@@ -3312,7 +3312,7 @@ static bool ggml_backend_hexagon_device_supports_op(ggml_backend_dev_t dev, cons\n break;\n \n case GGML_OP_GLU:\n- if ((ggml_get_glu_op(op) == GGML_GLU_OP_SWIGLU) /* || (ggml_get_glu_op(op) ==...
2025-12-17T21:38:21
huggingface/transformers
d5d87934c2294da88802fdef7dbd22b89f38f2be
81aabe72d3ab08b42e04d06a4e1f6b6fe92a5832
repo. consistency bot (#42575) * draft * draft * draft * draft * draft * draft * draft * draft * draft * draft * draft * fail to see the check * fail to see the check * fail to see the check * fail to see the check * fail to see the check * Apply style fixes * fail to see the check * fail to see the ...
[ { "path": ".github/workflows/pr-repo-consistency-bot.yml", "patch": "@@ -0,0 +1,218 @@\n+name: PR Repo. Consistency Bot\n+\n+on:\n+ issue_comment:\n+ types:\n+ - created\n+ branches-ignore:\n+ - main\n+concurrency:\n+ group: ${{ github.workflow }}-${{ github.event.issue.number }}-${{ start...
2025-12-03T16:57:21
denoland/deno
7c84b09f35bc13aec624921dd9507637d505527d
9e682be240965f2edab07363f2a72750bf250ac8
fix(ext/signals): prevent panic on FreeBSD (#32518) This PR fixes a panic in Deno on FreeBSD. The issue occurs because the signal dictionary for FreeBSD incorrectly includes SIGLIBRT (signal 33), which is not defined on FreeBSD, leading to an EINVAL error when attempting to bind or handle signals. Fixes #31087
[ { "path": "ext/signals/dict.rs", "patch": "@@ -104,8 +104,7 @@ signal_dict!(\n (29, \"SIGINFO\"),\n (30, \"SIGUSR1\"),\n (31, \"SIGUSR2\"),\n- (32, \"SIGTHR\"),\n- (33, \"SIGLIBRT\")\n+ (32, \"SIGTHR\")\n );\n \n #[cfg(target_os = \"openbsd\")]", "additions": 1, "deletions": 2, "languag...
2026-03-11T22:33:12
vuejs/vue
e396eb3445904f11232f2355f03e8356173d0e31
4e6d6379ee35f986a842c95964615dc27afa832f
fix(ref): allow ref key to be zero (#7676) prevents missing elements when :ref value is "0" fix #7669
[ { "path": "src/core/vdom/modules/ref.js", "patch": "@@ -1,6 +1,6 @@\n /* @flow */\n \n-import { remove } from 'shared/util'\n+import { remove, isDef } from 'shared/util'\n \n export default {\n create (_: any, vnode: VNodeWithData) {\n@@ -19,7 +19,7 @@ export default {\n \n export function registerRef (vn...
2018-02-21T15:21:51
ggml-org/llama.cpp
6ce3d857962e8183f1fbd7d5aeacf7534a427dc3
e85e9d7637268906d3fc75ec65bd2ef6ebea3a54
server: (webui) add --webui-config (#18028) * server/webui: add server-side WebUI config support Add CLI arguments --webui-config (inline JSON) and --webui-config-file (file path) to configure WebUI default settings from server side. Backend changes: - Parse JSON once in server_context::load_model() for performance ...
[ { "path": "common/arg.cpp", "patch": "@@ -2610,6 +2610,20 @@ common_params_context common_params_parser_init(common_params & params, llama_ex\n params.api_prefix = value;\n }\n ).set_examples({LLAMA_EXAMPLE_SERVER}).set_env(\"LLAMA_ARG_API_PREFIX\"));\n+ add_opt(common_arg(\n+ ...
2025-12-17T20:45:45
ollama/ollama
ed5fb088c4d7f57ea3ccf629e36ecb2e857e22eb
11d83386a5761b31f78e060054483c6f0c70dc7e
Fix target in gen_windows.ps1
[ { "path": "llm/generate/gen_windows.ps1", "patch": "@@ -26,7 +26,6 @@ function amdGPUs {\n $GPU_LIST -join ';'\n }\n \n-$script:cmakeTargets = @(\"ollama_llama_server\")\n \n function init_vars {\n if (!$script:SRC_DIR) {\n@@ -35,6 +34,9 @@ function init_vars {\n if (!$script:llamacppDir) {\n ...
2024-04-26T22:10:42
huggingface/transformers
81aabe72d3ab08b42e04d06a4e1f6b6fe92a5832
dd6cfdd3455f34986fe851375dbab0648f181a81
Align tie weights in Idefics (#42551) fix
[ { "path": "src/transformers/models/idefics/modeling_idefics.py", "patch": "@@ -1107,31 +1107,15 @@ def __init__(self, config, vision_model=None):\n bias=False,\n partially_freeze=config.freeze_lm_head,\n )\n+ if config.additional_vocab_size > 0:\n+ self._tie...
2025-12-03T16:56:05
denoland/deno
e0464f494ee846121c0dbbaf37e3bf270a9cb8e9
3249c98a5527fbaadc3549fe393fef4fcd803e16
fix(ext/fetch): retry on stale pooled HTTP/1.1 connections (#32566) ## Summary Fixes #31955 When a server shuts down and restarts on the same port, the fetch client's connection pool may still hold a stale keep-alive connection. Reusing it causes `connection closed before message completed` errors on every other req...
[ { "path": "ext/fetch/lib.rs", "patch": "@@ -1514,6 +1514,30 @@ fn is_error_retryable(err: &(dyn std::error::Error + 'static)) -> bool {\n }\n }\n \n+ // HTTP/1.1: The connection was closed before the message completed.\n+ // This happens when a pooled keep-alive connection is stale (e.g. the\n+ // ...
2026-03-11T20:08:18
ggml-org/llama.cpp
8dcc3662a292c14f003be2c465895d40c9460511
d37fc935059211454e9ad2e2a44e8ed78fd6d1ce
llama-fit-params: fix memory print (#18136)
[ { "path": "src/llama.cpp", "patch": "@@ -542,6 +542,7 @@ static void llama_params_fit_impl(\n } else {\n assert(ngl_per_device_high[id].n_layer == n_unassigned);\n ngl_per_device = ngl_per_device_high;\n+ mem = mem_high;\n ...
2025-12-17T20:10:03
huggingface/transformers
dd6cfdd3455f34986fe851375dbab0648f181a81
b8d5018e3fcd8f48a084526c983187d344e214f6
Add SDPA support for PatchTST model (#42465) * Add SDPA and Flash Attention support for PatchTST model - Add _supports_sdpa = True and _supports_flash_attn = True to PatchTSTPreTrainedModel - The existing PatchTSTAttention class already uses ALL_ATTENTION_FUNCTIONS to select the attention implementation based on co...
[ { "path": "src/transformers/models/patchtst/modeling_patchtst.py", "patch": "@@ -24,6 +24,7 @@\n \n from ... import initialization as init\n from ...activations import ACT2CLS\n+from ...integrations.deepspeed import is_deepspeed_zero3_enabled\n from ...modeling_flash_attention_utils import FlashAttentionKwa...
2025-12-03T16:17:14
ollama/ollama
f81f30811878ee3b59deaa2319eec489fafb39ef
7fea1ecdf68103cdc329b7f8c65904cd54c3e56d
fix gemma, command-r layer weights
[ { "path": "llm/memory.go", "patch": "@@ -102,10 +102,14 @@ func EstimateGPULayers(gpus []gpu.GpuInfo, ggml *GGML, projectors []string, opts\n \tlayers := ggml.Tensors().Layers()\n \n \tvar memoryLayerOutput uint64\n-\tfor k, v := range layers {\n-\t\tif k == \"output\" || k == \"output_norm\" {\n-\t\t\tmemo...
2024-04-26T22:00:54
denoland/deno
3249c98a5527fbaadc3549fe393fef4fcd803e16
b2ecb43e6229bad75498023c03c8b7f6abf73048
chore: fix windows build with latest msvc (#32644)
[ { "path": "Cargo.lock", "patch": "@@ -1870,7 +1870,7 @@ dependencies = [\n \"http-body 1.0.0\",\n \"http-body-util\",\n \"import_map\",\n- \"indexmap 2.9.0\",\n+ \"indexmap 2.12.0\",\n \"jsonc-parser 0.28.0\",\n \"jupyter-protocol\",\n \"keyring\",\n@@ -2064,7 +2064,7 @@ dependencies = [\n \"deno_med...
2026-03-11T19:08:23
ggml-org/llama.cpp
d37fc935059211454e9ad2e2a44e8ed78fd6d1ce
4470a0764a8940e2c867070d086a3e0259decfa0
webui: fix chat header width when sidebar is closed (#17981) * webui: fix chat header width when sidebar is closed * chore: add index.html.gz
[ { "path": "tools/server/webui/src/lib/components/app/chat/ChatScreen/ChatScreenHeader.svelte", "patch": "@@ -2,6 +2,9 @@\n \timport { Settings } from '@lucide/svelte';\n \timport { DialogChatSettings } from '$lib/components/app';\n \timport { Button } from '$lib/components/ui/button';\n+\timport { useSideba...
2025-12-17T19:05:45
denoland/deno
a395e3630d753bac2f8668a8e8e86198ac4b3c3d
cbdfed1dfe9a6664184fbb7e07608d841b8fafe7
refactor(ext/node): consolidate `node:fs` (part 7) (#32631) ## Summary Follow-up to #32588. Continues the node:fs consolidation series by inlining 4 more separate files into the main `fs.ts` polyfill: - `_fs/_fs_truncate.ts` → `truncate()`, `truncateSync()` - `_fs/_fs_utimes.ts` → `utimes()`, `utimesSync()`, `getVal...
[ { "path": "ext/node/lib.rs", "patch": "@@ -394,10 +394,6 @@ deno_core::extension!(deno_node,\n \"_fs/_fs_realpath.ts\",\n \"_fs/_fs_stat.ts\",\n \"_fs/_fs_statfs.ts\",\n- \"_fs/_fs_symlink.ts\",\n- \"_fs/_fs_truncate.ts\",\n- \"_fs/_fs_utimes.ts\",\n- \"_fs/_fs_watch.ts\",\n \"_n...
2026-03-11T13:32:48
ggml-org/llama.cpp
4470a0764a8940e2c867070d086a3e0259decfa0
4301e27319eadb8e129e91a0d9a4c96d9e2f6f15
ggml-hexagon: gelu operation (#17921) * feat: inital support for gelu using sigmoid approximation * snapshot: faster gelu using polynomial approximation * test: disable l2-block prefetch in polynomail approximation * Revert "test: disable l2-block prefetch in polynomail approximation" This reverts commit 72339994...
[ { "path": "ggml/src/ggml-hexagon/ggml-hexagon.cpp", "patch": "@@ -2161,8 +2161,14 @@ static bool ggml_hexagon_supported_activations(const struct ggml_hexagon_session\n }\n \n // src0, src1 & dst must be mapped to the same session\n- if (!hex_supported_buffer(sess, src0, src1, dst)) {\n- re...
2025-12-17T18:39:32
huggingface/transformers
ef780bf131fb506ca0095237f4d87e037c4b6234
7266f50bc8c3f071fd2fddb5eeb87280a029eabc
Add torch compile to CB (#42516) * transfer commit * Allow fullgraph and little fixes * fix when no measurements * CB is better at handling compile. Also can be benched. * Style * Add sumarrized by default * Doc and better CG logic * CG logic rollback * Update to Fa3 thx to Anton * style
[ { "path": "benchmark_v2/framework/benchmark_config.py", "patch": "@@ -5,6 +5,7 @@\n from functools import lru_cache\n from typing import Any\n \n+from transformers.generation.configuration_utils import CompileConfig\n from transformers.utils.import_utils import is_flash_attn_2_available, is_kernels_availabl...
2025-12-03T15:14:57
denoland/deno
cbdfed1dfe9a6664184fbb7e07608d841b8fafe7
7bcaaa2a124f505b87fa6f9c84e13abeb0a8c9de
fix(npm): handle read-only bin files when setting up node_modules/.bin (#32632) ## Summary - Some npm tarballs ship bin files with read+execute but **no write** permission (e.g. `@formatjs/cli@2.15.0` uses mode `555`). `make_executable_if_exists()` was opening these files with `O_RDWR` unconditionally, causing `EACCE...
[ { "path": "libs/npm_installer/bin_entries.rs", "patch": "@@ -382,11 +382,12 @@ fn make_executable_if_exists(\n path: &Path,\n ) -> Result<bool, std::io::Error> {\n let sys = sys.with_paths_in_errors();\n- let mut open_options = sys_traits::OpenOptions::new();\n- open_options.read = true;\n- open_opti...
2026-03-11T13:28:01
vuejs/vue
2d168f1ba11d77cb2f2caa3356de86e99aba8da6
8053eb1fba4202274efed11044679e011a1e525b
test(types): update TS version and fix types test (#7597)
[ { "path": "package-lock.json", "patch": "@@ -9580,9 +9580,9 @@\n \"dev\": true\n },\n \"typescript\": {\n- \"version\": \"2.6.2\",\n- \"resolved\": \"https://registry.npmjs.org/typescript/-/typescript-2.6.2.tgz\",\n- \"integrity\": \"sha1-PFtv1/beCRQmkCfwPAlGdY92c6Q=\",\n+ ...
2018-02-03T19:02:25
ggml-org/llama.cpp
a2c199e479bc739222c3ef24e563935dbaaf077c
15dd67d86984b3a61c40aacff367c37757172cc8
common: clarify instructions for bug reports (#18134)
[ { "path": "common/common.cpp", "patch": "@@ -1092,7 +1092,7 @@ common_init_result::common_init_result(common_params & params) :\n auto cparams = common_context_params_to_llama(params);\n \n if (params.fit_params) {\n- LOG_INF(\"%s: fitting params to device memory, to report bugs during this s...
2025-12-17T17:44:13
ollama/ollama
2a80f55e2ab0a94c3f087fc46523c14a70b0cd8e
36666c214270b7acf8d696a5c92f2fe33cfa14b8
Update windows.md (#3855) Fixed a typo
[ { "path": "docs/windows.md", "patch": "@@ -14,7 +14,7 @@ As this is a preview release, you should expect a few bugs here and there. If\n you run into a problem you can reach out on\r\n [Discord](https://discord.gg/ollama), or file an \r\n [issue](https://github.com/ollama/ollama/issues).\r\n-Logs will ofte...
2024-04-26T20:04:15
huggingface/transformers
7266f50bc8c3f071fd2fddb5eeb87280a029eabc
c0328af6d54d6ac367729427544f6e2772604248
[Quantization] Remove dequant fp8 config (#42596) * fix * style * up
[ { "path": "src/transformers/quantizers/quantizer_finegrained_fp8.py", "patch": "@@ -178,6 +178,10 @@ def _process_model_before_weight_loading(\n \n model.config.quantization_config = self.quantization_config\n \n+ def _process_model_after_weight_loading(self, model: \"PreTrainedModel\", **kwargs)...
2025-12-03T14:27:52
vuejs/vue
ced00b1dec8326a653cce225133927fe5b4a3109
4746256e553b19d4e37f264bf0c050fe0823d876
fix: fix config.productionTip fix #7565
[ { "path": "src/platforms/web/runtime/index.js", "patch": "@@ -44,32 +44,34 @@ Vue.prototype.$mount = function (\n \n // devtools global hook\n /* istanbul ignore next */\n-Vue.nextTick(() => {\n- if (config.devtools) {\n- if (devtools) {\n- devtools.emit('init', Vue)\n- } else if (\n- proce...
2018-01-31T07:53:25
denoland/deno
7bcaaa2a124f505b87fa6f9c84e13abeb0a8c9de
693607582b02998bcf81a47d61176fd4c6046f1f
fix(ext/node): handle signal 0 in child_process.kill() (#32294) ## Summary - `child_process.kill(0)` threw `Unknown signal: 0` because `toDenoSignal()` didn't recognize signal 0 - In POSIX, signal 0 checks if a process exists without sending a signal (`kill(pid, 0)`) - Now handles signal `0` (both numeric and string ...
[ { "path": "ext/node/polyfills/internal/child_process.ts", "patch": "@@ -576,6 +576,17 @@ export class ChildProcess extends EventEmitter {\n return this.killed;\n }\n \n+ // Signal 0 is a special case: it checks if the process exists\n+ // without sending a signal (POSIX kill(pid, 0)).\n+ ...
2026-03-11T13:15:29
ggml-org/llama.cpp
15dd67d86984b3a61c40aacff367c37757172cc8
bde461de8cff5a1e7b693b29e98db8db421ddd42
model: fix GLM-ASR-Nano-2512 load error (#18130) (#18142)
[ { "path": "tools/mtmd/mtmd.cpp", "patch": "@@ -329,6 +329,7 @@ struct mtmd_context {\n case PROJECTOR_TYPE_QWEN25O:\n case PROJECTOR_TYPE_ULTRAVOX:\n case PROJECTOR_TYPE_VOXTRAL:\n+ case PROJECTOR_TYPE_GLMA:\n audio_preproc = std::make_unique<mt...
2025-12-17T15:34:35
ollama/ollama
85801317d1d09379f85512f84ad37ab6761a5225
2ed0d65948195aecff0d63566d8087b4b64c6a61
Fix clip log import
[ { "path": "llm/patches/02-clip-log.diff", "patch": "@@ -0,0 +1,12 @@\n+diff --git a/examples/llava/clip.cpp b/examples/llava/clip.cpp\n+index e431c7f7..f077e688 100644\n+--- a/examples/llava/clip.cpp\n++++ b/examples/llava/clip.cpp\n+@@ -3,6 +3,7 @@\n+ // I'll gradually clean and extend it\n+ // Note: Even ...
2024-04-26T16:43:46
huggingface/transformers
c0328af6d54d6ac367729427544f6e2772604248
17c7c496000a5208370489d2cc4dec593b6dd66f
Fix the FA2 logic in the longcat_flash model (#42549) * Matching FA2 naming under kernels * make style * convert model * Follow the comments
[ { "path": "src/transformers/models/longcat_flash/modeling_longcat_flash.py", "patch": "@@ -431,7 +431,7 @@ def forward(\n cache_kwargs = {\"sin\": sin, \"cos\": cos, \"cache_position\": cache_position}\n key_states, value_states = past_key_values.update(key_states, value_states, self...
2025-12-03T14:21:48
vuejs/vue
eaf6b54a389dd6e316df932b3fe3a5d6579e2499
04da767d10038d355b67778173ccf66ea74211c5
chore: fix another link
[ { "path": "BACKERS.md", "patch": "@@ -41,7 +41,7 @@ Funds donated via Patreon goes directly to support Evan You's full-time work on\n </tr><tr></tr>\n <tr>\n <td align=\"center\" valign=\"middle\">\n- <a href=\"https://datacamp.com/careers?utm_source=vuejs&utm_medium=sidebar\" target=\"...
2018-01-12T14:47:00
ggml-org/llama.cpp
669696e00d60e7ffa0223ee61242318a51f33a79
982060fadcf2469363efeaab113d7f81eaff4bf6
ggml-cpu: ARM64: repack version of q8_0 (dotprod and i8mm) (#18096) * wip: skeleton for q8_0 repack * q8_0 repack GEMV implementations * GEMM implementations * Formatting * Fixed format consistency of repack gemm and gemv declarations * gemv and gemm generic location consistent with declarations * Removed non-co...
[ { "path": "ggml/src/ggml-cpu/arch-fallback.h", "patch": "@@ -43,6 +43,8 @@\n #define ggml_gemv_q2_K_8x8_q8_K_generic ggml_gemv_q2_K_8x8_q8_K\n #define ggml_gemv_iq4_nl_4x4_q8_0_generic ggml_gemv_iq4_nl_4x4_q8_0\n #define ggml_gemv_iq4_nl_8x8_q8_0_generic ggml_gemv_iq4_nl_8x8_q8_0\n+#define ggml_gemv_q8_0_4x...
2025-12-17T11:39:13
denoland/deno
26972e65e0c74699f313f3b1181adb6243e6aeaf
9bb8552cf9d6bed5233007d135c4b0523522e672
fix(ext/node): fix TLS JSStreamSocket resource leak causing process hang (#32325) Closes https://github.com/denoland/deno/issues/30556 When `tls.connect({ socket })` receives a non-`net.Socket` stream (e.g. mssql/tedious wrapping TDS framing in a Duplex via native-duplexpair), Deno creates a `JSStreamSocket` to br...
[ { "path": "ext/node/ops/tls.rs", "patch": "@@ -336,6 +336,8 @@ struct JSDuplexResource {\n readable: Arc<Mutex<tokio::sync::mpsc::Receiver<Bytes>>>,\n writable: tokio::sync::mpsc::Sender<Bytes>,\n read_buffer: Arc<Mutex<VecDeque<Bytes>>>,\n+ closed: AtomicBool,\n+ close_notify: tokio::sync::Notify,\...
2026-03-11T13:10:26
huggingface/transformers
17c7c496000a5208370489d2cc4dec593b6dd66f
130dc4710f6a7f459171c48a3f41c0da5bbb2166
Fix mixed torch.Tensor and DTensor in generate when use FSDP2 + LoRA (#42436) * Fix mixed torch.Tensor and DTensor error by registering fsdp forward for trainer.mode.generate * Apply fsdp forward register when fsdp is enabled --------- Co-authored-by: yiminzme <yiminzme@gmail.com> Co-authored-by: Ferdinand Mom <474...
[ { "path": "src/transformers/trainer.py", "patch": "@@ -2338,6 +2338,8 @@ def _inner_training_loop(\n \n if self.is_fsdp_enabled:\n self.model = self.model_wrapped = model\n+ # Fix `got mixed torch.Tensor and DTensor` error in model.generate() for FSDP2 with LoRA\n+ ...
2025-12-03T14:05:36
vuejs/vue
04da767d10038d355b67778173ccf66ea74211c5
f0e9dc44672c6b68d919f26f5ccc49af21385a59
chore: fix sponsor link
[ { "path": "BACKERS.md", "patch": "@@ -72,8 +72,8 @@ Funds donated via Patreon goes directly to support Evan You's full-time work on\n </a>\n </td>\n <td align=\"center\" valign=\"middle\">\n- <a href=\"https://starter.someline.com/\" target=\"_blank\">\n- <img width=\"148...
2018-01-12T14:45:34
ggml-org/llama.cpp
982060fadcf2469363efeaab113d7f81eaff4bf6
6853bee680a128231ff35d0152a7d38d42afa3ad
model: fix LFM2_MOE missing tensors (#18132)
[ { "path": "src/llama-arch.cpp", "patch": "@@ -2055,7 +2055,7 @@ static std::set<llm_tensor> llm_get_tensor_names(llm_arch arch) {\n LLM_TENSOR_SHORTCONV_INPROJ,\n LLM_TENSOR_SHORTCONV_OUTPROJ,\n LLM_TENSOR_TOKEN_EMBD,\n- LLM_TENSOR_OUTPUT_NORM,\...
2025-12-17T11:17:11
huggingface/transformers
130dc4710f6a7f459171c48a3f41c0da5bbb2166
8ba286ba4e7025c07451bc9a6d22a422ff5934a3
fix : cast into floats AFTER all assignments (#42587) Co-authored-by: marconaguib <marco.naguib@aphp.fr>
[ { "path": "src/transformers/data/data_collator.py", "patch": "@@ -711,9 +711,6 @@ def __post_init__(self):\n if self.random_replace_prob < 0 or self.random_replace_prob > 1:\n raise ValueError(\"random_replace_prob should be between 0 and 1.\")\n \n- self.mask_replace_prob = float...
2025-12-03T13:40:39
denoland/deno
9bb8552cf9d6bed5233007d135c4b0523522e672
0238181e8a95be0885da93bbc99997988d07bff0
fix(ext/node): handle cargo features during publish (#32636) hot fix for v2.7.5 release that broke during `cargo publish` step.
[ { "path": "ext/node/ops/fs.rs", "patch": "@@ -9,6 +9,7 @@ use std::rc::Rc;\n use deno_core::OpState;\n use deno_core::ResourceId;\n use deno_core::op2;\n+#[cfg(feature = \"sync_fs\")]\n use deno_core::unsync::spawn_blocking;\n use deno_fs::FileSystemRc;\n use deno_fs::FsFileType;\n@@ -20,8 +21,24 @@ use den...
2026-03-11T12:38:55
vuejs/vue
5191f13472d1fc37bdd601079970201fde6bf13e
05299610ea3e89ddbcfe4d8ede0c298223766423
fix(transition): should not add transition class when cancelled (#7391) fix #7390
[ { "path": "src/platforms/web/runtime/modules/transition.js", "patch": "@@ -149,13 +149,15 @@ export function enter (vnode: VNodeWithData, toggleDisplay: ?() => void) {\n addTransitionClass(el, startClass)\n addTransitionClass(el, activeClass)\n nextFrame(() => {\n- addTransitionClass(el, to...
2018-01-05T22:45:59
ggml-org/llama.cpp
487674fbb31d20ad13f9dd7f81da00f650c34e25
acec774ef619c81d4c59adca0e548b3ba035554f
common: fix --override-kv to support comma-separated values (#18056) * common: fix --override-kv to support comma-separated values * Update common/arg.cpp Co-authored-by: Xuan-Son Nguyen <thichthat@gmail.com> * common: deprecate repeated arguments, suggest comma-separated values * common: add comma escape support ...
[ { "path": "common/arg.cpp", "patch": "@@ -420,6 +420,8 @@ static bool common_params_parse_ex(int argc, char ** argv, common_params_context\n }\n };\n \n+ std::set<std::string> seen_args;\n+\n for (int i = 1; i < argc; i++) {\n const std::string arg_prefix = \"--\";\n \n@@ -430,6 +...
2025-12-17T09:36:23
huggingface/transformers
8ba286ba4e7025c07451bc9a6d22a422ff5934a3
cba7ae865b16f85b6008aebb220461a8863075f4
fix: correct typos in code comments (#42577) Fix three typos found in code comments: - 'avaoid' → 'avoid' in modeling_utils.py - 'weigth' → 'weight' in trainer_utils.py - 'Templace' → 'Template' in convert_slow_tokenizer.py These typos appeared in TODO comments and inline documentation.
[ { "path": "src/transformers/convert_slow_tokenizer.py", "patch": "@@ -1171,7 +1171,7 @@ def converted(self) -> Tokenizer:\n )\n tokenizer.decoder = decoders.ByteLevel()\n \n- # Hack to have a ByteLevel and TemplaceProcessor\n+ # Hack to have a ByteLevel and TemplateProcessor\n ...
2025-12-03T13:33:49