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
48126d6a93e9809f9964b0ab65c9acd51baf41f0
b910888867ac91f245ce87f68e389259a9a69fe2
fix(ext/node): support setImmediate promisify.custom (#31920) - Add dedicated `setImmediatePromise` function in `timers.ts` - Add `promisify.custom` property to `setImmediate` that returns `setImmediatePromise` - Update `promises.setImmediate` to use `setImmediatePromise` directly - Bump `node_shim` to 0.3.0 (includes...
[ { "path": "Cargo.lock", "patch": "@@ -6356,9 +6356,9 @@ dependencies = [\n \n [[package]]\n name = \"node_shim\"\n-version = \"0.2.0\"\n+version = \"0.3.0\"\n source = \"registry+https://github.com/rust-lang/crates.io-index\"\n-checksum = \"4c0c15839a684a17f9f07e62811a8989ba204e8b0f011859d5d0f73a4667d82b\"\...
2026-01-22T22:44:31
huggingface/transformers
313afcc468d5e2500bb1eeb210ece0151166a1ee
7bba4d12025b2e76aa0962a0dbacd4bf944d19a9
[chat template] update when "push_to_hub" (#39815) * update templates push to hub * rvert jinja suffix and move it to processor file
[ { "path": "src/transformers/processing_utils.py", "patch": "@@ -776,6 +776,7 @@ def save_pretrained(self, save_directory, push_to_hub: bool = False, legacy_seri\n Additional key word arguments passed along to the [`~utils.PushToHubMixin.push_to_hub`] method.\n \"\"\"\n use_au...
2025-10-15T13:49:59
ggml-org/llama.cpp
d00cbea63c671cd85a57adaa50abf60b3b87d86f
8328fd4bae76fc027f8ca0e9a05acd3788dabe3f
server : host-memory prompt caching (#16391) * minor : code style * server : fix prompt similarity calculation * server : initial host-memory prompt caching * cont * server : refactor * cont * cont : make the server task of the slot const * cont : minor [no ci] * server : cache prompts and checkpoints only for...
[ { "path": "common/arg.cpp", "patch": "@@ -1935,6 +1935,14 @@ common_params_context common_params_parser_init(common_params & params, llama_ex\n params.n_ctx_checkpoints = value;\n }\n ).set_env(\"LLAMA_ARG_CTX_CHECKPOINTS\").set_examples({LLAMA_EXAMPLE_SERVER}));\n+ add_opt(common...
2025-10-09T15:54:51
ollama/ollama
0625e805f09e0ba7ea04448842621ec0e829d725
c38ec5befbb9ab50074e3794ec867b30c845cb0e
fix model name not matching
[ { "path": "cmd/cmd.go", "patch": "@@ -127,13 +127,9 @@ func RunHandler(cmd *cobra.Command, args []string) error {\n \t\treturn err\n \t}\n \n-\tmodelName, modelTag, ok := strings.Cut(args[0], \":\")\n-\tif !ok {\n-\t\tmodelTag = \"latest\"\n-\t}\n-\n+\tcanonicalModelPath := server.ParseModelPath(args[0])\n ...
2023-09-27T02:49:55
vuejs/vue
ec7fca8495528bda169c44ac660747bf2a8000a8
8bf5af8edf966e6fad8f260234519cf40ab0cde6
Fix multiple attr in select with undefined value (#4859) Fix #4855
[ { "path": "src/platforms/web/runtime/node-ops.js", "patch": "@@ -7,7 +7,8 @@ export function createElement (tagName: string, vnode: VNode): Element {\n if (tagName !== 'select') {\n return elm\n }\n- if (vnode.data && vnode.data.attrs && 'multiple' in vnode.data.attrs) {\n+ // false or null will r...
2017-02-06T16:58:36
ggml-org/llama.cpp
8328fd4bae76fc027f8ca0e9a05acd3788dabe3f
56b4795842d852152222ca7a2d4304008facf1b9
No markdown in cot (#16483) * fix: let the model think in plaintext * chore: npm run format + npm run build
[ { "path": "tools/server/webui/src/lib/components/app/chat/ChatMessages/ChatMessageThinkingBlock.svelte", "patch": "@@ -4,7 +4,6 @@\n \timport * as Collapsible from '$lib/components/ui/collapsible/index.js';\n \timport { buttonVariants } from '$lib/components/ui/button/index.js';\n \timport { Card } from '$l...
2025-10-09T15:36:29
huggingface/transformers
7bba4d12025b2e76aa0962a0dbacd4bf944d19a9
ab925343775847c224af9653cb141efd25846167
Fix video processing channel format (#41603) fix
[ { "path": "src/transformers/video_processing_utils.py", "patch": "@@ -57,12 +57,12 @@\n VideoInput,\n VideoMetadata,\n group_videos_by_shape,\n+ infer_channel_dimension_format,\n is_valid_video,\n load_video,\n make_batched_metadata,\n make_batched_videos,\n reorder_videos...
2025-10-15T13:48:01
denoland/deno
b910888867ac91f245ce87f68e389259a9a69fe2
b9b87a6bfb61a5d8855ee1578620ac3a9393b6bb
fix(ext/node): use primordials in `ext/node/polyfills/tls.ts` (#31816) Towards #24236. Replaces JS builtins with equivalent primordials.
[ { "path": "ext/node/polyfills/tls.ts", "patch": "@@ -1,9 +1,6 @@\n // Copyright 2018-2026 the Deno authors. MIT license.\n // Copyright Joyent and Node contributors. All rights reserved. MIT license.\n \n-// TODO(petamoriken): enable prefer-primordials for node polyfills\n-// deno-lint-ignore-file prefer-pr...
2026-01-22T22:44:08
ollama/ollama
d5ad41dd7b95434d313695d48951e01ce0558414
d294a11bc9d7e7094870698173e152c32758e63c
fix path for wsl user
[ { "path": "scripts/install.sh", "patch": "@@ -82,6 +82,7 @@ Group=ollama\n Restart=always\n RestartSec=3\n Environment=\"HOME=/usr/share/ollama\"\n+Environment=\"PATH=$PATH\"\n \n [Install]\n WantedBy=default.target", "additions": 1, "deletions": 0, "language": "Shell Script" } ]
2023-09-26T00:56:25
vuejs/vue
8bf5af8edf966e6fad8f260234519cf40ab0cde6
e853d1994dd87b14fd1577e1fc301c88333944c8
fix `addClass` utlil bug (#4863) * fix addClass util bug `<transition-group move-class=“ value with leading or trailing whitespace ” />` * fix addClass util bug `<transition-group move-class=“ value with leading or trailing whitespace ” />` * fix addClass util bug `<transition-group move-class=“ value w...
[ { "path": "src/platforms/web/runtime/class-util.js", "patch": "@@ -6,7 +6,7 @@\n */\n export function addClass (el: Element, cls: ?string) {\n /* istanbul ignore if */\n- if (!cls || !cls.trim()) {\n+ if (!cls || !(cls = cls.trim())) {\n return\n }\n \n@@ -31,7 +31,7 @@ export function addClass (...
2017-02-06T16:53:04
huggingface/transformers
ab925343775847c224af9653cb141efd25846167
56a727dde5c29d902d58d984d346affc223ae241
enable sdpa enable gqa logic for Ascend NPU (#41601) * enable gqa logic for Ascend NPU * remove redundant comments * fix comments about Ascend NPU --------- Co-authored-by: Anton Vlasjuk <73884904+vasqu@users.noreply.github.com>
[ { "path": "src/transformers/integrations/sdpa_attention.py", "patch": "@@ -29,19 +29,15 @@ def repeat_kv(hidden_states: torch.Tensor, n_rep: int) -> torch.Tensor:\n \n def use_gqa_in_sdpa(attention_mask: Optional[torch.Tensor], key: torch.Tensor) -> bool:\n # GQA can only be used under the following con...
2025-10-15T13:45:28
ggml-org/llama.cpp
aa4711d369b0d4adb6802b98ad6ea362f838710e
d80d6d2400b8faa80654c723cb5bdf9fc8f4db06
CANN: Improve ACL graph matching (#16166) * CANN: improve ACL graph matching Record `ne` and `nb` information for src tensors and include them in the graph matching check. This enhances the robustness of ACL graph matching by preventing incorrect matches when src tensors share the same data address but differ in shap...
[ { "path": "ggml/src/ggml-cann/common.h", "patch": "@@ -341,11 +341,18 @@ class cann_task_queue {\n \n #ifdef USE_ACL_GRAPH\n struct ggml_graph_node_properties {\n+ // dst tensor\n void * node_address;\n- ggml_op node_op;\n int64_t ne[GGML_MAX_DIMS];\n size_t nb[GGML_MAX_DIMS];\n+\n+ // ...
2025-10-09T07:50:25
denoland/deno
98557599cbde32d99806a57182e50881af5e2538
2a91db0259480b4c91afc82a4758d63b9fe76faf
feat: instrument cron with OTEL (#31916) This PR adds OpenTelemetry tracing support to `Deno.cron()` handlers. When OTEL tracing is enabled, each cron handler invocation is wrapped in a span named `"deno.cron"` with `deno.cron.name` and `deno.cron.schedule` attributes. The span is properly entered so that any child sp...
[ { "path": "ext/cron/01_cron.ts", "patch": "@@ -1,15 +1,19 @@\n // Copyright 2018-2026 the Deno authors. MIT license.\n \n import { core, internals, primordials } from \"ext:core/mod.js\";\n-const {\n- isPromise,\n-} = core;\n import { op_cron_create, op_cron_next } from \"ext:core/ops\";\n const {\n Arra...
2026-01-22T13:55:20
vuejs/vue
b80f8ba5ab74d46bdebc68589b97d281197e1144
068095579b1349e883f508fd69793886ea415d5a
only map ssr stack trace if error is indeed an Error
[ { "path": "src/server/create-bundle-renderer.js", "patch": "@@ -50,7 +50,9 @@ export function createBundleRendererCreator (createRenderer: () => Renderer) {\n runInVm(entry, files, context).then(app => {\n renderer.renderToString(app, cb)\n }).catch(err => {\n- rewriteErro...
2017-01-30T18:42:18
huggingface/transformers
56a727dde5c29d902d58d984d346affc223ae241
dc6fdeb7052d23a7efde1e7dc793c17e2cfe98b0
Add fast path for bidirectional mask creation to fix regression (#41586) * fixed performance regression * also fixed the older_torch function * Update src/transformers/masking_utils.py Co-authored-by: Anton Vlasjuk <73884904+vasqu@users.noreply.github.com> * fix * more general * fix slicing * fix data dependent...
[ { "path": "src/transformers/masking_utils.py", "patch": "@@ -405,6 +405,16 @@ def sdpa_mask_recent_torch(\n if allow_is_bidirectional_skip and _ignore_bidirectional_mask_sdpa(padding_mask):\n return None\n \n+ # vmap can incur performance issues as reported in #41566 for bidirectional mask as...
2025-10-15T13:30:39
ggml-org/llama.cpp
b2602137557b2b28a39e03612717d85ead9a6f5a
e08db4259521de493b7aeb49dadf29ebd1ee966a
[SYCL] refactor soft_max, add soft_max_back (#16472) * refactor to support soft_max_ext * fix error and support soft_max_back * rm unused functions * fix format issue --------- Co-authored-by: Zhang Jianyu <zhang.jianyu@outlook.com>
[ { "path": "ggml/src/ggml-sycl/common.hpp", "patch": "@@ -197,6 +197,7 @@ struct sycl_device_info {\n int cc; // compute capability\n // int nsm; // number of streaming multiprocessors\n // size_t smpb; // max. shared memory per block\n+ si...
2025-10-09T07:25:11
ollama/ollama
d0854bf1e663f77a537c5fcccdd97577fa02b686
83964632550830d9569dfc8dce5c640fe3007aa1
fix dkms on debian
[ { "path": "scripts/install.sh", "patch": "@@ -204,7 +204,7 @@ if ! lsmod | grep -q nvidia; then\n debian|ubuntu) $SUDO apt-get -y install linux-headers-$KERNEL_RELEASE ;;\n esac\n \n- NVIDIA_CUDA_VERSION=$(dkms status | awk -F: '/added/ { print $1 }')\n+ NVIDIA_CUDA_VERSION=$($SUDO dkms st...
2023-09-25T19:57:21
denoland/deno
2a91db0259480b4c91afc82a4758d63b9fe76faf
f321e88fdf071c49c117a4d4018dee0a2def4214
fix(ext/node): inspector.open() and inspector.close() (#31898) This commit adds proper support for `node:inspector` APIs: - `inspector.open()` - `inspector.close()` Also `process.features.inspector` value was changed to `true`. As a result a bunch of Node compat tests related to inspector started failing - turns out...
[ { "path": "Cargo.lock", "patch": "@@ -1885,9 +1885,9 @@ dependencies = [\n \n [[package]]\n name = \"deno_core\"\n-version = \"0.380.0\"\n+version = \"0.380.1\"\n source = \"registry+https://github.com/rust-lang/crates.io-index\"\n-checksum = \"669603e9d217e82488f22f89632831a7bc879f7cae9827037d44f22c27148ba...
2026-01-22T10:55:15
huggingface/transformers
dc6fdeb7052d23a7efde1e7dc793c17e2cfe98b0
3953b6544021657fd3105b2b69b81bab811bf044
Update a dataset reop link (#41618) fix Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>
[ { "path": "tests/models/kosmos2_5/test_processor_kosmos2_5.py", "patch": "@@ -299,7 +299,7 @@ def test_structured_kwargs_nested_from_dict(self):\n \n @require_torch\n def test_full_processor(self):\n- url = url_to_local_path(\"https://huggingface.co/kirp/kosmos2_5/resolve/main/receipt_00008.p...
2025-10-15T12:41:38
ggml-org/llama.cpp
e08db4259521de493b7aeb49dadf29ebd1ee966a
12bbc3fa50b6df03318a4451c9a2210200a0b28d
model: EmbeddingGemma Adding Support for SentenceTransformers Dense Modules (#16367) * model: EmbeddingGemma sentence-transformers dense linear projections support * model: add support for EmbeddingGemma SentenceTransformers dense linear projections Adding support for the Dense modules used in EmbeddingGemma models....
[ { "path": "convert_hf_to_gguf.py", "patch": "@@ -93,13 +93,15 @@ class ModelBase:\n # Mistral format specifics\n is_mistral_format: bool = False\n disable_mistral_community_chat_template: bool = False\n+ sentence_transformers_dense_modules: bool = False\n \n def __init__(self, dir_model: ...
2025-10-09T06:39:18
denoland/deno
f321e88fdf071c49c117a4d4018dee0a2def4214
1556f14651887bf85e02ad71a7376426f0076967
fix(ext/node): validate fd in tty.isatty and enable pseudo-tty tests (#31892) Closes https://github.com/denoland/deno/issues/29876 - tty.isatty() now returns false for negative and non-integer fd values - Added run_in_pty() to run pseudo-tty tests in a real terminal - Skip pseudo-tty tests on Windows CI where PTY i...
[ { "path": "ext/node/polyfills/tty.js", "patch": "@@ -17,7 +17,7 @@ import * as io from \"ext:deno_io/12_io.js\";\n \n // Returns true when the given numeric fd is associated with a TTY and false otherwise.\n function isatty(fd) {\n- if (typeof fd !== \"number\") {\n+ if (typeof fd !== \"number\" || fd >> ...
2026-01-22T10:34:06
huggingface/transformers
3953b6544021657fd3105b2b69b81bab811bf044
96d245a83da0c5b154b98ef6559f9981cc1aefc2
Reinstate early CUDA init fix (#41617) * Reinstate early CUDA init fix Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com> * Delay import further Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com> --------- Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github...
[ { "path": "src/transformers/generation/continuous_batching/continuous_api.py", "patch": "@@ -30,7 +30,6 @@\n from ...configuration_utils import PretrainedConfig\n from ...generation.configuration_utils import GenerationConfig\n from ...generation.logits_process import LogitsProcessor\n-from ...integrations....
2025-10-15T12:41:10
ggml-org/llama.cpp
12bbc3fa50b6df03318a4451c9a2210200a0b28d
9d0882840e6c3fb62965d03af0e22880ea90e012
refactor: centralize CoT parsing in backend for streaming mode (#16394) * refactor: unify reasoning handling via backend reasoning_content, drop frontend tag parsing - Updated the chat message component to surface backend-supplied reasoning via message.thinking while showing the raw assistant content without inline t...
[ { "path": "common/arg.cpp", "patch": "@@ -3432,7 +3432,8 @@ common_params_context common_params_parser_init(common_params & params, llama_ex\n {\"--reasoning-format\"}, \"FORMAT\",\n \"controls whether thought tags are allowed and/or extracted from the response, and in which format they're r...
2025-10-08T20:18:41
ollama/ollama
a027bbf4d7eb73c7448393f84f6181d0ab791a97
ed94a3dd029575a4035e7edf265154c52c82e4d8
fix dkms install
[ { "path": "scripts/install.sh", "patch": "@@ -138,7 +138,6 @@ install_cuda_driver_yum() {\n esac\n \n status 'Installing CUDA driver...'\n- $SUDO $PACKAGE_MANAGER -y update\n \n if [ \"$1\" = 'centos' ] || [ \"$1$2\" = 'rhel7' ]; then\n $SUDO $PACKAGE_MANAGER -y install nvidia-driver-...
2023-09-25T18:28:21
vuejs/vue
d57f9428140993d8b04a13e54ef06e195d31f195
d14bd641430d936867a5d289d513a37917a0deb6
fixed typo on line 15 (#4829)
[ { "path": "src/core/vdom/helpers/normalize-children.js", "patch": "@@ -12,7 +12,7 @@ import VNode, { createTextVNode } from 'core/vdom/vnode'\n \n // 1. When the children contains components - because a functional component\n // may return an Array instead of a single root. In this case, just a simple\n-// ...
2017-02-01T15:10:37
denoland/deno
7046dc85c45ab809267d715dc72250ead11705a7
f7e4aa3d95613561101a84b5c587c50033b80923
fix(ext/node): improve child_process CLI argument parsing (#31825) Closes https://github.com/denoland/deno/issues/26567 Replace incomplete JavaScript-based Node CLI flag handling with a comprehensive Rust parser for better `node:child_process` compatibility. --------- Co-authored-by: Bartek Iwańczuk <biwanczuk@gmai...
[ { "path": "Cargo.lock", "patch": "@@ -2509,6 +2509,7 @@ dependencies = [\n \"md4\",\n \"nix 0.27.1\",\n \"node_resolver\",\n+ \"node_shim\",\n \"num-bigint\",\n \"num-bigint-dig\",\n \"num-integer\",\n@@ -6352,6 +6353,15 @@ dependencies = [\n \"url\",\n ]\n \n+[[package]]\n+name = \"node_shim\"\n+ver...
2026-01-22T00:45:13
ggml-org/llama.cpp
d2ee056e1df0fdf646270fb5621e9f92084b59a7
b2c08c9ec4cd59ed88c1ebdd94f109af5ccc978e
server : fix cancel pending task (#16467) Co-authored-by: DevAI <DevAI@gmail.com>
[ { "path": "tools/server/server.cpp", "patch": "@@ -1937,7 +1937,7 @@ struct server_queue {\n void cleanup_pending_task(int id_target) {\n // no need lock because this is called exclusively by post()\n auto rm_func = [id_target](const server_task & task) {\n- return task.id_tar...
2025-10-08T08:20:18
huggingface/transformers
bb0c3af995385fd070083610c98130fe9341c87a
70e871959c3ced65ee4804a55fb27b37876db2bf
More markdown file fixes (#41599) * Format markdown files Signed-off-by: Yuanyuan Chen <cyyever@outlook.com> * Format markdown files Signed-off-by: Yuanyuan Chen <cyyever@outlook.com> * Format markdown files Signed-off-by: Yuanyuan Chen <cyyever@outlook.com> --------- Signed-off-by: Yuanyuan Chen <cyyever@outlo...
[ { "path": "docs/source/en/index.md", "patch": "@@ -36,8 +36,6 @@ Explore the [Hub](https://huggingface.com/) today to find a model and use Transf\n \n Explore the [Models Timeline](./models_timeline) to discover the latest text, vision, audio and multimodal model architectures in Transformers.\n \n-\n-\n ##...
2025-10-15T12:29:27
vuejs/vue
e7083d09f1eec1d8853b65720f1a8c6e0f22836e
6cbee6b286f705a86bf5502c49f764fe2bd2d30e
fix scoped slots with dynamic slot names + force update for child components with scoped slots (fix #4779)
[ { "path": "flow/component.js", "patch": "@@ -104,6 +104,8 @@ declare interface Component {\n _b: (data: any, value: any, asProp?: boolean) => VNodeData;\n // check custom keyCode\n _k: (eventKeyCode: number, key: string, builtInAlias: number | Array<number> | void) => boolean;\n+ // resolve scoped sl...
2017-01-24T16:04:02
ggml-org/llama.cpp
b2c08c9ec4cd59ed88c1ebdd94f109af5ccc978e
7fdd16b432d247121fe5fe7b21f8805f85266c85
metal : mark FA blocks (#16372) * metal : better unroll in the FA kernels * metal : index FA blocks * tests : restore [no ci] * metal : prevent division by zero in FA kernels * metal : fix -INF detection logic
[ { "path": "ggml/src/ggml-metal/ggml-metal-device.cpp", "patch": "@@ -959,7 +959,53 @@ ggml_metal_pipeline_t ggml_metal_library_get_pipeline_flash_attn_ext_pad(\n //ggml_metal_cv_set_int32(cv, ns20, FC_FLASH_ATTN_EXT_PAD + 21);\n //ggml_metal_cv_set_int32(cv, nsg, FC_FLASH_ATTN_EXT_PAD + 22);\n //ggml...
2025-10-08T07:57:53
denoland/deno
f7e4aa3d95613561101a84b5c587c50033b80923
0be96f999828fb2d60cc26aca5166eca20493d45
fix(doc): non-determinism in `deno doc --lint` (#31911)
[ { "path": "Cargo.lock", "patch": "@@ -1991,9 +1991,9 @@ dependencies = [\n \n [[package]]\n name = \"deno_doc\"\n-version = \"0.189.0\"\n+version = \"0.189.1\"\n source = \"registry+https://github.com/rust-lang/crates.io-index\"\n-checksum = \"5d82f89d6417f786a32d310e3287306ad4c0763a53a83b8323f86b9ac471c71e...
2026-01-21T23:33:02
huggingface/transformers
70e871959c3ced65ee4804a55fb27b37876db2bf
c4210796e073f1a85b8140b7cecba0d5d6d0c71c
Fix trainer simple tests (#41449) * fix * fix ray * train to tune * breaking changes wrt generation config * Fix ! * fix * fix * fix deepspeed ! * fix * fix * fix * improve logic * revert and fix * revert comment * oups * revert change * fix * style * typo in comment --------- Co-authored-by: Cyril...
[ { "path": "src/transformers/integrations/integration_utils.py", "patch": "@@ -302,7 +302,7 @@ def run_hp_search_ray(trainer, n_trials: int, direction: str, **kwargs) -> BestR\n for more options\n \"\"\"\n import ray\n- import ray.train\n+ import ray.tune\n \n def _objective(tri...
2025-10-15T12:09:00
vuejs/vue
90a455c95ceb5d38e5a26b172e958232c25265d2
379695c9126fcc00ab7af1a663bd97f949b03ad1
fix replaced component root nodes losing parent scopeId (fix #4774)
[ { "path": "src/core/vdom/patch.js", "patch": "@@ -260,9 +260,14 @@ export function createPatchFunction (backend) {\n // of going through the normal attribute patching process.\n function setScope (vnode) {\n let i\n- if (isDef(i = vnode.context) && isDef(i = i.$options._scopeId)) {\n- nodeOp...
2017-01-23T16:42:59
denoland/deno
b0d0578d62df3481d39b2e05244443b8852ff0e5
f8160bfd93d7f3c88512d7fa55e82cc98d84a9c9
fix(dx): always spawn subprocess for dx (#31897) Fixes #31692. Fixes #31874. previously I tried to avoid this, but it messes up the argv0. So just spawn a process. Since we're going through native code (as opposed to a bash wrapper) it doesn't add too much time (I measured) though it does hurt perf a bit.
[ { "path": "cli/main.rs", "patch": "@@ -257,9 +257,9 @@ async fn run_subcommand(\n DenoSubcommand::Repl(repl_flags) => {\n spawn_subcommand(async move { tools::repl::run(flags, repl_flags).await })\n }\n- DenoSubcommand::X(x_flags) => spawn_subcommand(async move {\n- tools::x::run(flags...
2026-01-21T19:46:01
ggml-org/llama.cpp
74b8fc17f92ada295a648e3c5eb28f46bca7d892
aeaf8a36f06b5810f5ae4bbefe26edb33925cf5e
ggml webgpu: profiling, CI updates, reworking of command submission (#16452) * Add profiling * More detailed profiling * Rework command submission to avoid global locks * Update wait handling * try new method of waiting on futures * Add serializing of command submission in some cases * Add new pool for timestamp...
[ { "path": ".github/workflows/build.yml", "patch": "@@ -444,8 +444,8 @@ jobs:\n # This is using llvmpipe and runs slower than other backends\n ctest -L main --verbose --timeout 4200\n \n- ubuntu-22-cmake-webgpu:\n- runs-on: ubuntu-22.04\n+ ubuntu-24-cmake-webgpu:\n+ runs-on: ubunt...
2025-10-07T20:48:56
huggingface/transformers
fcd1ccdb78e7cc28a880e815ee883549f334a5ae
2b2c20f3155f843bef95d340920bf20a8b87bac7
[`Docs`] Fix changed references (#41614) * fix * fix * other ln
[ { "path": "docs/source/en/models.md", "patch": "@@ -139,10 +139,10 @@ with tempfile.TemporaryDirectory() as tmp_dir:\n new_model = AutoModel.from_pretrained(tmp_dir)\n ```\n \n-Sharded checkpoints can also be directly loaded with [`~transformers.modeling_utils.load_sharded_checkpoint`].\n+Sharded checkp...
2025-10-15T11:59:13
vuejs/vue
769c4dc2032251323c8f61ad8eba2c26c615a618
b1e66b489144ec31ad3d4f3298e3809f00d6c4e6
ensure mutating extended constructor options does not affect parent (fix #4767)
[ { "path": "src/core/instance/state.js", "patch": "@@ -39,12 +39,14 @@ const isReservedProp = { key: 1, ref: 1, slot: 1 }\n \n function initProps (vm: Component, props: Object) {\n const propsData = vm.$options.propsData || {}\n- const keys = vm.$options._propKeys = Object.keys(props)\n+ // cache prop ke...
2017-01-22T18:01:00
ollama/ollama
c65edb15068b763fbd9e8e1afd014ab53064ad14
1605af32ec8d21798ca7599535863c5a3f7a19a7
fix linux installer warning logs (#588)
[ { "path": "scripts/install.sh", "patch": "@@ -6,6 +6,7 @@ set -eu\n \n status() { echo \">>> $*\" >&2; }\n error() { echo \"ERROR $*\"; exit 1; }\n+warning() { echo \"WARNING: $*\"; }\n \n TEMP_DIR=$(mktemp -d)\n cleanup() { rm -rf $TEMP_DIR; }\n@@ -95,7 +96,7 @@ fi\n check_gpu() {\n case $1 in\n ...
2023-09-25T15:22:56
ggml-org/llama.cpp
4e0388aa8a1f4ef1065701dc9c3947aea0a1b9a5
ef4c5b87ea2556ff8ca99cca3abdf48bdbca22f2
webui : added download action (#13552) (#16282) * webui : added download action (#13552) * webui : import and export (for all conversations) * webui : fixed download-format, import of one conversation * webui : add ExportedConversations type for chat import/export * feat: Update naming & order * chore: Linting *...
[ { "path": "tools/server/webui/src/lib/components/app/chat/ChatSidebar/ChatSidebarActions.svelte", "patch": "@@ -1,8 +1,9 @@\n <script lang=\"ts\">\n-\timport { Search, SquarePen, X } from '@lucide/svelte';\n+\timport { Search, SquarePen, X, Download, Upload } from '@lucide/svelte';\n \timport { KeyboardShor...
2025-10-07T09:11:08
huggingface/transformers
2b2c20f3155f843bef95d340920bf20a8b87bac7
e2122c4bcb74d942bb93c11dcb55aafc4c7fdf23
Update issue template (#41573) * update * fix
[ { "path": ".github/ISSUE_TEMPLATE/bug-report.yml", "patch": "@@ -48,18 +48,17 @@ body:\n - continuous batching: @remi-or @ArthurZucker @McPatate\n - pipelines: @Rocketknight1\n - tokenizers: @ArthurZucker and @itazap\n- - trainer: @zach-huggingface @SunMarc\n+ ...
2025-10-15T11:54:37
denoland/deno
f8160bfd93d7f3c88512d7fa55e82cc98d84a9c9
d2a9b8c77a467396dd23008bc10a4c33df5f3cb9
fix(publish): do not include tsconfig imports in publish graph (#31907)
[ { "path": "cli/factory.rs", "patch": "@@ -812,6 +812,7 @@ impl CliFactory {\n self.caches()?.clone(),\n self.cjs_tracker()?.clone(),\n cli_options.clone(),\n+ self.compiler_options_resolver()?.clone(),\n self.file_fetcher()?.clone(),\n s...
2026-01-21T19:45:07
vuejs/vue
938fa4efcc9bf6232bf5ace5920398dc2e128ac9
9e38abca3e6ababe0b06dfc9838a90e06ff5ea88
fix(sfc): component contains '<' only (#4753) * fix(sfc): component contains '<' only * update based on reviews
[ { "path": "src/compiler/parser/html-parser.js", "patch": "@@ -123,7 +123,7 @@ export function parseHTML (html, options) {\n }\n \n let text, rest, next\n- if (textEnd > 0) {\n+ if (textEnd >= 0) {\n rest = html.slice(textEnd)\n while (\n !endTag.test(rest) &&"...
2017-01-20T03:23:20
ollama/ollama
b5614f3ebcb0fd9167ead2627b066d50555a780d
8b2ba9cab822b80e873c8b7554504eebfaf434c2
fix end-of-line issue with the new prompt (#582)
[ { "path": "cmd/cmd.go", "patch": "@@ -33,15 +33,22 @@ import (\n )\n \n type Painter struct {\n-\tHideHint bool\n+\tIsMultiLine bool\n }\n \n func (p Painter) Paint(line []rune, _ int) []rune {\n \ttermType := os.Getenv(\"TERM\")\n-\tif termType == \"xterm-256color\" && len(line) == 0 && !p.HideHint {\n-\t\...
2023-09-24T00:20:30
denoland/deno
7e6483145f1a11111c68a443c405a8b7b38edad0
6c22c7393931779622adbd1cf726efbef073444e
fix(ext/node): prevent derefencing symlinks on `fs.rm` (#31886)
[ { "path": "ext/node/ops/winerror.rs", "patch": "@@ -30,7 +30,6 @@ use deno_core::op2;\n #[string]\n pub fn op_node_sys_to_uv_error(err: i32) -> String {\n let uv_err = match err {\n- ERROR_ACCESS_DENIED => \"EACCES\",\n ERROR_NOACCESS => \"EACCES\",\n WSAEACCES => \"EACCES\",\n ERROR_CANT_A...
2026-01-20T23:56:47
ggml-org/llama.cpp
ef4c5b87ea2556ff8ca99cca3abdf48bdbca22f2
c61ae20d05bd4fdd8551311325a2845336449426
presets : fix pooling param for embedding models (#16455)
[ { "path": "common/arg.cpp", "patch": "@@ -3859,7 +3859,6 @@ common_params_context common_params_parser_init(common_params & params, llama_ex\n [](common_params & params) {\n params.model.hf_repo = \"ggml-org/bge-small-en-v1.5-Q8_0-GGUF\";\n params.model.hf_file = \"bge-small-...
2025-10-07T07:32:32
vuejs/vue
9e38abca3e6ababe0b06dfc9838a90e06ff5ea88
466e849c85e21521c9ede360f44b0b72be2acb22
Support select multiple binding (fix #4755) (#4756) * support select multiple binding * improve select onchange handle * update style
[ { "path": "src/platforms/web/compiler/directives/model.js", "patch": "@@ -163,13 +163,14 @@ function genSelect (\n }\n \n const number = modifiers && modifiers.number\n- const assignment = `Array.prototype.filter` +\n+ const selectedVal = `Array.prototype.filter` +\n `.call($event.target.options,f...
2017-01-19T18:36:19
ollama/ollama
e29662ab5c7cbab869bd2b55326179a552c57b15
cbc40aa9967d27a58f87442018f2d889782d1980
fix minor install script issues on debian
[ { "path": "scripts/install.sh", "patch": "@@ -30,12 +30,14 @@ case \"$(uname -m)\" in\n *) error \"Unsupported architecture: $ARCH\" ;;\n esac\n \n-SUDO_CMD=\n+SUDO=\n if [ \"$(id -u)\" -ne 0 ]; then\n # Running as root, no need for sudo\n if ! command -v sudo >/dev/null; then\n error \"...
2023-09-23T14:25:47
ggml-org/llama.cpp
3df2244df40c67dfd6ad548b40ccc507a066af2b
c08002a1988348403a5fd59b1fa3de3a10a6f92f
llama : add --no-host to disable host buffers (#16310) * implement --no-host to disable host buffer * fix equal_mparams * move no-host enumeration order together with other model params --------- Co-authored-by: slaren <slarengh@gmail.com>
[ { "path": "common/arg.cpp", "patch": "@@ -2584,6 +2584,13 @@ common_params_context common_params_parser_init(common_params & params, llama_ex\n params.no_extra_bufts = true;\n }\n ).set_env(\"LLAMA_ARG_NO_REPACK\"));\n+ add_opt(common_arg(\n+ {\"--no-host\"},\n+ \"by...
2025-10-06T17:55:53
vuejs/vue
458671b732c709999cbd1f0fa37d38215cab839b
d01104e1242c870de12592da88d2c4ffd0c9c0c2
handle include/exclude changing to non-valid types (fix #4749)
[ { "path": "src/core/components/keep-alive.js", "patch": "@@ -12,9 +12,10 @@ function getComponentName (opts: ?VNodeComponentOptions): ?string {\n function matches (pattern: string | RegExp, name: string): boolean {\n if (typeof pattern === 'string') {\n return pattern.split(',').indexOf(name) > -1\n- ...
2017-01-19T15:13:55
ollama/ollama
e20362e0d5fef9e87e34c003ead3feafa9bba24c
c928ceb927bf27ec00fd0d424300362c86927c8d
fix multi line input in `ollama run`
[ { "path": "cmd/cmd.go", "patch": "@@ -32,11 +32,13 @@ import (\n \t\"github.com/jmorganca/ollama/version\"\n )\n \n-type Painter struct{}\n+type Painter struct {\n+\tHideHint bool\n+}\n \n-func (p Painter) Paint(line []rune, l int) []rune {\n+func (p Painter) Paint(line []rune, _ int) []rune {\n \ttermType ...
2023-09-23T03:49:35
huggingface/transformers
e89cef66252379cf19254d247ac9bb1bd2ab487f
26b7f668500281fe96c0ae4a9adcb603a60f0427
fix some case failures lead by "`torch.compile` recompiled part of th… (#41558) * fix some case failures lead by "`torch.compile` recompiled part of the forward pass" in xpu Signed-off-by: Wang, Yi A <yi.a.wang@intel.com> * update comment Signed-off-by: Wang, Yi A <yi.a.wang@intel.com> --------- Signed-off-by: Wa...
[ { "path": "src/transformers/masking_utils.py", "patch": "@@ -839,7 +839,8 @@ def create_causal_mask(\n # Do not allow skip if we are compiling (this is to match BC)\n # TODO: cyril -> probably revisit and remove this, but a lot of tests rely on it\n if _is_torch_xpu_available:\n- allow_is...
2025-10-15T10:45:29
ggml-org/llama.cpp
c08002a1988348403a5fd59b1fa3de3a10a6f92f
3a002afafa8e06555f11aed88b02d055d8a166f3
chat : Granite Docling stopping (#16438) * fix: Fix duplicate fake image before token on first slice Branch: GraniteDoclingStopping Signed-off-by: Gabe Goodhart <ghart@us.ibm.com> * fix: Use double-newline before overview image Branch: GraniteDoclingStopping Signed-off-by: Gabe Goodhart <ghart@us.ibm.com> * fix:...
[ { "path": "src/llama-chat.cpp", "patch": "@@ -590,7 +590,7 @@ int32_t llm_chat_apply_template(\n ss << message->content << \"<|end_of_text|>\\n\";\n }\n if (add_ass) {\n- ss << \"<|start_of_role|>assistant<|end_of_role|>\\n\";\n+ ss << \"<|start_of_role|>ass...
2025-10-06T16:59:40
vuejs/vue
55c87aa2c5fbb40f94d8b1cc0f14bf07cf5a1885
1ec501ebbfe7cc51e6f84d3143c6e21eebb34699
fix multi-select test case bug detection in IE9
[ { "path": "test/unit/features/directives/model-select.spec.js", "patch": "@@ -6,10 +6,11 @@ import { looseEqual } from 'shared/util'\n function hasMultiSelectBug () {\n var s = document.createElement('select')\n s.setAttribute('multiple', '')\n- s.innerHTML = '<option>1</option>'\n- s.options[0].selec...
2017-01-16T23:39:06
ggml-org/llama.cpp
a23b9bdbd3b64ce172f9962249f432d01aea7437
04e632a4aab8e6bfff0f8bc216b36ceb1e199ff9
ggml : fix unaligned access in AMX code (#16315)
[ { "path": "ggml/src/ggml-cpu/amx/amx.cpp", "patch": "@@ -149,6 +149,7 @@ class extra_buffer_type : ggml::cpu::extra_buffer_type {\n if (op->op == GGML_OP_MUL_MAT && is_contiguous_2d(op->src[0]) && // src0 must be contiguous\n is_contiguous_2d(op->src[1]) && ...
2025-10-06T13:05:27
huggingface/transformers
5db730786d87bde8397a207bd84ae0014054fc37
13a35a5057cbdc34b6c93a26d4e57987cbdd205c
[device_map] Accelerate loading by computing device_map much faster (#41548) * start * add the important fix * continue * big cleanup * type hints * add method * fix typehints * typehints * fix * oupsi * remove space * improve function * CI
[ { "path": "src/transformers/integrations/accelerate.py", "patch": "@@ -12,20 +12,23 @@\n # See the License for the specific language governing permissions and\n # limitations under the License.\n \"\"\"\n-Since, https://github.com/huggingface/transformers/pull/36963, loading is always performed with models ...
2025-10-15T09:18:57
vuejs/vue
1ec501ebbfe7cc51e6f84d3143c6e21eebb34699
85d6f115bb9adbc7fbba57d4fbb445daaad79b52
fix raf bind in IE9
[ { "path": "src/platforms/web/runtime/transition-util.js", "patch": "@@ -28,7 +28,10 @@ if (hasTransition) {\n }\n \n // binding to window is necessary to make hot reload work in IE in strict mode\n-const raf = (inBrowser && window.requestAnimationFrame.bind(window)) || setTimeout\n+const raf = inBrowser && ...
2017-01-16T23:10:53
denoland/deno
f8df593f34bbe12af4720dbbd5520ded399655f1
7f553d3e3adad912074d5483b84615cc3549a166
fix(ext/web): export PerformanceObserver and PerformanceObserverEntryList (#31887) Follow up https://github.com/denoland/deno/pull/31875 that exposes Web APIs on the global scope.
[ { "path": "runtime/js/98_global_scope_shared.js", "patch": "@@ -83,6 +83,10 @@ const windowOrWorkerGlobalScope = {\n PerformanceEntry: core.propNonEnumerable(performance.PerformanceEntry),\n PerformanceMark: core.propNonEnumerable(performance.PerformanceMark),\n PerformanceMeasure: core.propNonEnumera...
2026-01-20T13:33:36
ollama/ollama
88897a90e49e79b1c5733527d87f28d45bac03f4
9df31c351865a5e4ebc7cca47ad6f888c85268f1
fix ipv6 parse ip
[ { "path": "cmd/cmd.go", "patch": "@@ -650,7 +650,7 @@ func RunServer(cmd *cobra.Command, _ []string) error {\n \thost, port, err := net.SplitHostPort(os.Getenv(\"OLLAMA_HOST\"))\n \tif err != nil {\n \t\thost, port = \"127.0.0.1\", \"11434\"\n-\t\tif ip := net.ParseIP(os.Getenv(\"OLLAMA_HOST\")); ip != nil ...
2023-09-22T17:40:54
ggml-org/llama.cpp
04e632a4aab8e6bfff0f8bc216b36ceb1e199ff9
a80ff183abe4e5a76316257ffa597da41b3b6fa0
ci : remove missing reranker model files (#16444) This commit removes jina-reranker-v1-tiny-en model files that are no longer present on Hugging Face. The motivation for this that it clears up the CI logs from 404 errors which can be a little confusing when looking at the logs the first time. Refs: https://github.co...
[ { "path": "ci/run.sh", "patch": "@@ -512,12 +512,7 @@ function gg_run_rerank_tiny {\n gg_wget models-mnt/rerank-tiny/ https://huggingface.co/jinaai/jina-reranker-v1-tiny-en/raw/main/tokenizer_config.json\n gg_wget models-mnt/rerank-tiny/ https://huggingface.co/jinaai/jina-reranker-v1-tiny-en/raw/mai...
2025-10-06T12:56:59
huggingface/transformers
13a35a5057cbdc34b6c93a26d4e57987cbdd205c
94df0e65602922be2831b3faa457a2bde78b936b
Enable non-streaming mode in `transformers serve` (#41446) * Enable non-streaming in transformers serve Remove typos Remove typos Remove typos * Fix tests * Arthur review
[ { "path": "src/transformers/commands/serving.py", "patch": "@@ -26,7 +26,7 @@\n import time\n import uuid\n from argparse import ArgumentParser, Namespace\n-from collections.abc import AsyncGenerator, Generator, Iterable\n+from collections.abc import Generator, Iterable\n from contextlib import asynccontext...
2025-10-15T07:37:26
vuejs/vue
85d6f115bb9adbc7fbba57d4fbb445daaad79b52
c66991d961469f70b9649813e8bf5eb2f231fead
ensure instance-unique keys for <transition> child nodes during pending node removal (fix #4702)
[ { "path": "src/platforms/web/runtime/components/transition.js", "patch": "@@ -127,9 +127,13 @@ export default {\n return placeholder(h, rawChild)\n }\n \n- const key = child.key = child.key == null || child.isStatic\n- ? `__v${child.tag + this._uid}__`\n- : child.key\n+ // ensure a...
2017-01-16T23:00:34
denoland/deno
7f553d3e3adad912074d5483b84615cc3549a166
8812083d1cc27ba8320f1b545c118284cf574520
fix(ext/sqlite): double free of Session after database closed (#31830)
[ { "path": "ext/node/ops/sqlite/session.rs", "patch": "@@ -112,6 +112,9 @@ impl Session {\n }\n \n self.freed.set(true);\n+ if self.db.borrow().is_none() {\n+ return Ok(());\n+ }\n // Safety: `self.inner` is a valid session. double free is\n // prevented by `freed` flag.\n unsa...
2026-01-20T12:31:48
ggml-org/llama.cpp
a80ff183abe4e5a76316257ffa597da41b3b6fa0
1d49ca37594fb49db6aa9518ba7c512e5ccd0108
ggml-cpu : fix leftover handling in ggml_vec_scale_f32 for SVE (#16443) This commit updates the leftover handling in ggml_vec_scale_f32. The motivation for this is that the code currently incorrectly assumes there would be fewer than ggml_f32_epr leftover elements. However, since the main loop processes 2*ggml_f32_ep...
[ { "path": "ggml/src/ggml-cpu/vec.h", "patch": "@@ -654,11 +654,11 @@ inline static void ggml_vec_scale_f32(const int n, float * y, const float v) {\n }\n // leftovers\n // maximum number of leftover elements will be less that ggml_f32_epr. Apply predicated svmad on available elemen...
2025-10-06T12:17:12
huggingface/transformers
94df0e65602922be2831b3faa457a2bde78b936b
9e4199ede396f136b3dff1e918816fcc3a65f0a0
Benchmark overhaul (#41408) * Big refactor, still classes to move around and script to re-complexify * Move to streamer, isolate benches, propagate num tokens * Some refacto * Added compile mode to name * Re-order * Move to dt_tokens * Better format * Fix and disable use_cache by default * Fixed compile and SD...
[ { "path": ".github/workflows/benchmark.yml", "patch": "@@ -1,10 +1,7 @@\n name: Self-hosted runner (benchmark)\r\n \r\n on:\r\n- push:\r\n- branches: [main]\r\n- pull_request:\r\n- types: [ opened, labeled, reopened, synchronize ]\r\n+ workflow_dispatch:\r\n \r\n concurrency:\r\n group: ${{ githu...
2025-10-14T19:41:43
vuejs/vue
c66991d961469f70b9649813e8bf5eb2f231fead
a23b22e91d9c26552b40939d2396bd2330225412
Fix hot reload in IE in strict mode (#4725) Following a fix for v1 at #4274 for issue #4465
[ { "path": "src/platforms/web/runtime/transition-util.js", "patch": "@@ -27,7 +27,8 @@ if (hasTransition) {\n }\n }\n \n-const raf = (inBrowser && window.requestAnimationFrame) || setTimeout\n+// binding to window is necessary to make hot reload work in IE in strict mode\n+const raf = (inBrowser && window....
2017-01-16T21:45:24
denoland/deno
821d36129fe292cff8ee8a00c48a5a6d113985cd
88341cadf6bd924e1369ecd7f9cbd636ae122bb0
fix(ext/crypto): add support for SHA3 (#31668) Fix #31631
[ { "path": "Cargo.lock", "patch": "@@ -1961,6 +1961,7 @@ dependencies = [\n \"ed448-goldilocks\",\n \"elliptic-curve\",\n \"num-traits\",\n+ \"ocb3\",\n \"once_cell\",\n \"p256\",\n \"p384\",\n@@ -1971,6 +1972,7 @@ dependencies = [\n \"serde_bytes\",\n \"sha1\",\n \"sha2\",\n+ \"sha3\",\n \"signatu...
2026-01-19T17:23:04
ggml-org/llama.cpp
ca71fb9b368e3db96e028f80c4c9df6b6b370edd
35266573b968e1c947b367782fb4b3eddbb4f3c0
model : Granite docling + Idefics3 preprocessing (SmolVLM) (#16206) * feat: Add granite-docling conversion using trillion pretokenizer Branch: gabe-l-hart/GraniteDocling Signed-off-by: Gabe Goodhart <ghart@us.ibm.com> * feat: Add granite-docling vocab pre enum Branch: gabe-l-hart/GraniteDocling Signed-off-by: Gab...
[ { "path": "convert_hf_to_gguf.py", "patch": "@@ -891,6 +891,9 @@ def get_vocab_base_pre(self, tokenizer) -> str:\n if chkhsh == \"9b1be57e70d20d9501b2b3186e792d81181ae36ada3903c26f9fea418cf87206\":\n # ref: https://huggingface.co/inclusionAI/LLaDA-MoE-7B-A1B-Base\n res = \"ll...
2025-10-05T12:57:47
ollama/ollama
c98669436742f79f8ffaa796ff94ee2c7f17201a
058d0cd04b6f73bab1ac478a7483a469adc32355
fix HEAD / request HEAD request should respond like their GET counterparts except without a response body.
[ { "path": "server/routes.go", "patch": "@@ -543,12 +543,11 @@ func Serve(ln net.Listener, allowOrigins []string) error {\n \t\t},\n \t)\n \n-\tr.GET(\"/\", func(c *gin.Context) {\n-\t\tc.String(http.StatusOK, \"Ollama is running\")\n-\t})\n-\tr.HEAD(\"/\", func(c *gin.Context) {\n-\t\tc.Status(http.StatusOK...
2023-09-21T23:35:38
huggingface/transformers
9e4199ede396f136b3dff1e918816fcc3a65f0a0
4c8d293599c73b2bec56cdaec39366f330038e2d
Gemma3 fixes (#41572) * Multiple device error fix * FA2 equivalence fix * Move the train fwd in cfg test * Style * Added comment * Made the comment more clear
[ { "path": "src/transformers/models/gemma3/modeling_gemma3.py", "patch": "@@ -798,7 +798,7 @@ def create_causal_mask_mapping(\n is_previous_image = nn.functional.pad(is_image, (1, 0), value=0)[:, :-1]\n new_image_start = is_image & ~is_previous_image\n image_group_ids = torch.cumsum(n...
2025-10-14T16:33:27
vuejs/vue
fe6a26bb9c689acba0ee2b5619c17eeb9426a871
ee6ad6af0e7d39f7f7b22bb5931045d748e03ca7
keep-alive: prune cache on include/exclude change, also avoid firing deactivate for already inactive components (fix #4633)
[ { "path": "src/core/components/keep-alive.js", "patch": "@@ -5,6 +5,10 @@ import { getFirstComponentChild } from 'core/vdom/helpers/index'\n \n const patternTypes = [String, RegExp]\n \n+function getComponentName (opts: ?VNodeComponentOptions): ?string {\n+ return opts && (opts.Ctor.options.name || opts.ta...
2017-01-13T23:54:35
denoland/deno
88341cadf6bd924e1369ecd7f9cbd636ae122bb0
c66d1254b34ba91d803294f1f349e8e2a6b6567c
fix(ext/node): implement PerformanceObserver (#31875) Fixes #31723 --------- Co-authored-by: Felipe Cardozo <fraifelipe@gmail.com>
[ { "path": "cli/tsc/dts/node/perf_hooks.d.cts", "patch": "@@ -559,6 +559,11 @@ declare module \"perf_hooks\" {\n */\n class PerformanceObserver extends AsyncResource {\n constructor(callback: PerformanceObserverCallback);\n+ /**\n+ * Returns a list of the entry types supported...
2026-01-19T16:11:17
ggml-org/llama.cpp
35266573b968e1c947b367782fb4b3eddbb4f3c0
86df2c9ae4f2f1ee63d2558a9dc797b98524639b
ggml webgpu: actually add softmax, fix rms_norm offset (#16400) * implement soft_max * Fix soft_max data race * Temporary fix, wait on each submit
[ { "path": "ggml/src/ggml-webgpu/ggml-webgpu.cpp", "patch": "@@ -424,6 +424,7 @@ static void ggml_backend_webgpu_build_and_enqueue(webgpu_context &\n ctx->staged_param_bufs.push_back(params_bufs);\n if (ctx->staged_command_bufs.size() == WEBGPU_COMMAND_SUBMIT_BATCH_SIZE) {\n ggml_...
2025-10-05T03:59:31
huggingface/transformers
4c8d293599c73b2bec56cdaec39366f330038e2d
a99b1be3c73173509f71c261834d93d3d5e30ef1
Fix typsetting and content of llm_tutorial_optimization.md (#41172) * Fix typsetting of llm_tutorial_optimization Signed-off-by: Yuanyuan Chen <cyyever@outlook.com> * Fix errors Signed-off-by: Yuanyuan Chen <cyyever@outlook.com> --------- Signed-off-by: Yuanyuan Chen <cyyever@outlook.com>
[ { "path": "docs/source/en/llm_tutorial_optimization.md", "patch": "@@ -16,18 +16,18 @@ rendered properly in your Markdown viewer.\n Large Language Models (LLMs) such as GPT3/4, [Falcon](https://huggingface.co/tiiuae/falcon-40b), and [Llama](https://huggingface.co/meta-llama/Llama-2-70b-hf) are rapidly advan...
2025-10-14T15:40:26
ollama/ollama
1fabba474b3a1362b85da0e088dcd3967f39151b
765770efdbb8db463dfc359ae1f9653af908c779
refactor default allow origins this should be less error prone
[ { "path": "server/routes.go", "patch": "@@ -499,23 +499,25 @@ func CopyModelHandler(c *gin.Context) {\n \t}\n }\n \n-func Serve(ln net.Listener, origins []string) error {\n+var defaultAllowOrigins = []string{\n+\t\"localhost\",\n+\t\"127.0.0.1\",\n+\t\"0.0.0.0\",\n+}\n+\n+func Serve(ln net.Listener, allowOr...
2023-09-21T16:42:16
vuejs/vue
ee6ad6af0e7d39f7f7b22bb5931045d748e03ca7
1a1952ba81cf3580660122ff14254c10c3fda191
fix v-on .once on multiple elements (fix #4655)
[ { "path": "src/platforms/web/runtime/modules/events.js", "patch": "@@ -4,11 +4,17 @@ import { updateListeners } from 'core/vdom/helpers/index'\n \n let target: HTMLElement\n \n-function add (event: string, handler: Function, once: boolean, capture: boolean) {\n+function add (\n+ event: string,\n+ handler:...
2017-01-13T17:14:23
denoland/deno
c66d1254b34ba91d803294f1f349e8e2a6b6567c
2d77536bd96129b2e2a86a7f28c6d0b08d53d2c5
chore: fix flaky websocket unit test (#31883) This test previously assumed frame websocket frame arrived in the same read as HTTP headers. Now it just reads until we have everything.
[ { "path": "tests/unit/websocket_test.ts", "patch": "@@ -1125,12 +1125,24 @@ Deno.test(\"WebSocket close with reason but no code doesn't send 1005\", async ()\n );\n \n const response = new Uint8Array(2048);\n- const n = await conn.read(response);\n+ let totalRead = 0;\n+\n+ let n = await conn.read(re...
2026-01-19T09:39:31
ggml-org/llama.cpp
86df2c9ae4f2f1ee63d2558a9dc797b98524639b
f39283960b58a92ecc0c72567711318b20e22b55
vulkan: use a more appropriate amount of threads when generating shaders (#16418) * use a more flexible amount of threads * fix windows compile and 0 thread case * nominmax
[ { "path": "ggml/src/ggml-vulkan/vulkan-shaders/vulkan-shaders-gen.cpp", "patch": "@@ -1,5 +1,3 @@\n-\n-\n #include <iostream>\n #include <fstream>\n #include <sstream>\n@@ -22,6 +20,7 @@\n #include <sys/types.h>\n \n #ifdef _WIN32\n+ #define NOMINMAX\n #include <windows.h>\n #include <direct.h> /...
2025-10-04T20:04:27
huggingface/transformers
a99b1be3c73173509f71c261834d93d3d5e30ef1
82cae9eb520a2f852f79d71337548f9813c2aaa8
Revert some breaking changes bnb (#41581) fix
[ { "path": "src/transformers/quantizers/auto.py", "patch": "@@ -121,17 +121,15 @@ class AutoQuantizationConfig:\n @classmethod\n def from_dict(cls, quantization_config_dict: dict):\n quant_method = quantization_config_dict.get(\"quant_method\")\n- if quant_method is None:\n+ # W...
2025-10-14T14:28:16
vuejs/vue
1a1952ba81cf3580660122ff14254c10c3fda191
08bd81f8c0bd39816b5b509c5132a12188b412d7
handle single -> array & array -> single event handler patching (fix #4650)
[ { "path": "src/core/vdom/helpers/update-listeners.js", "patch": "@@ -9,10 +9,12 @@ export function updateListeners (\n remove: Function,\n vm: Component\n ) {\n- let name, cur, old, fn, event, capture, once\n+ let name, cur, old, curIsArray, oldIsArray, fn, event, capture, once\n for (name in on) {\...
2017-01-13T16:49:49
huggingface/transformers
82cae9eb520a2f852f79d71337548f9813c2aaa8
4fad35ee4ac3ebc63cb9781beca0706a3d43875e
Add __iter__ to DynamicCache (#41569) * Add __iter__ to DynamicCache * Fix tests that use ddp init
[ { "path": "src/transformers/cache_utils.py", "patch": "@@ -176,6 +176,11 @@ def __init__(self, sliding_window: int):\n super().__init__()\n self.sliding_window = sliding_window\n self.cumulative_length = 0\n+ self._sliding_window_tensor = torch.tensor(self.sliding_window, dtyp...
2025-10-14T14:16:32
ggml-org/llama.cpp
898acba6816ad23b6a9491347d30e7570bffadfd
e29acf74fea996014380d59d31aa504ae8964258
rpc : add support for multiple devices (#16276) * rpc : add support for multiple devices Allow rpc-server to expose multiple devices from a single endpoint. Change RPC protocol to include device identifier where needed. closes: #15210 * fixes * use ggml_backend_reg_t * address review comments * fix llama-bench b...
[ { "path": "common/arg.cpp", "patch": "@@ -1615,18 +1615,14 @@ static void add_rpc_devices(const std::string & servers) {\n if (!rpc_reg) {\n throw std::invalid_argument(\"failed to find RPC backend\");\n }\n- typedef ggml_backend_dev_t (*ggml_backend_rpc_add_device_t)(const char * endpoin...
2025-10-04T09:49:16
ollama/ollama
9297ff8330381e0e1e42db3af2e674413a33ba37
ee4fd16f2c02a643e70c5393f7bb27cfda58671f
fix OLLAMA_HOST parsing for ip6
[ { "path": "cmd/cmd.go", "patch": "@@ -655,28 +655,19 @@ func generateBatch(cmd *cobra.Command, model string) error {\n }\n \n func RunServer(cmd *cobra.Command, _ []string) error {\n-\thost, port := \"127.0.0.1\", \"11434\"\n-\n-\tparts := strings.Split(os.Getenv(\"OLLAMA_HOST\"), \":\")\n-\tif ip := net.Pa...
2023-09-21T00:49:48
denoland/deno
aa8652c55945826259375d5feddc8d1d434ac005
021b3b9e99c2605bde70b7b137d9a5c88ecc8379
fix(publish): include `npm:@types/` package resolution in published packages (#31855)
[ { "path": "Cargo.lock", "patch": "@@ -9194,9 +9194,9 @@ dependencies = [\n \n [[package]]\n name = \"sys_traits\"\n-version = \"0.1.17\"\n+version = \"0.1.22\"\n source = \"registry+https://github.com/rust-lang/crates.io-index\"\n-checksum = \"4f74a2c95f72e36fa6bd04a40d15623a9904bab1cc2fa6c6135b09d774a65088...
2026-01-18T19:03:52
vuejs/vue
08bd81f8c0bd39816b5b509c5132a12188b412d7
f4f3a015309ce361aa9ccc675ff2bb6ad1f36dd6
fix #4530 with smaller change scope
[ { "path": "flow/compiler.js", "patch": "@@ -6,7 +6,7 @@ declare type CompilerOptions = {\n directives?: { [key: string]: Function }; // platform specific directives\n isUnaryTag?: (tag: string) => ?boolean; // check if a tag is unary for the platform\n isReservedTag?: (tag: string) => ?boolean; // che...
2017-01-13T02:41:23
huggingface/transformers
8fe4db53994cdccf7629284add65655e6ce73af4
c620c38bb04c80bb12350bec739c298918b5675d
[kernels] rm mra kernels (#41507) * fix modeling * remove kernel * fix style
[ { "path": "src/transformers/kernels/mra/cuda_kernel.cu", "patch": "@@ -1,383 +0,0 @@\n-#include \"cuda_kernel.h\"\n-\n-//////////////////////////////////////////////////////////////////////////////////////////////////\n-////////////////////////////////////////////////////////////////////////////////////////...
2025-10-14T11:34:04
ggml-org/llama.cpp
e29acf74fea996014380d59d31aa504ae8964258
128d522c04286e019666bd6ee4d18e3fbf8772e2
vulkan : incremental shader builds (#16341) * vulkan (DRAFT): split shader generation by GLSL source file, to improve incremental build times * support dep-files so shaders are recompiled if their included files change * rename shader files which are used as "headers" to use .glsl extension * move glslc extension de...
[ { "path": "ggml/src/ggml-vulkan/CMakeLists.txt", "patch": "@@ -1,5 +1,6 @@\n cmake_minimum_required(VERSION 3.19)\n cmake_policy(SET CMP0114 NEW)\n+cmake_policy(SET CMP0116 NEW)\n \n find_package(Vulkan COMPONENTS glslc REQUIRED)\n \n@@ -54,25 +55,25 @@ if (Vulkan_FOUND)\n # Test all shader extensions\n...
2025-10-04T09:42:56
vuejs/vue
ab0a2259e0ef3f4ebdc8d0d4b929fae3e3f579d8
4e830ba3c3c92e78bf3ffbea82a583865deaa52a
Revert "also bind static special attrs as props (fix #4530)" This reverts commit b3ebfef91d72b301b85bce07c4ffad4f59bc73e2.
[ { "path": "src/compiler/parser/index.js", "patch": "@@ -477,15 +477,6 @@ function processAttrs (el) {\n }\n }\n addAttr(el, name, JSON.stringify(value))\n- // #4530 also bind special attributes as props even if they are static\n- // so that patches between dynamic/static are co...
2017-01-13T01:58:32
denoland/deno
021b3b9e99c2605bde70b7b137d9a5c88ecc8379
28184e1c5bec78836accdd6e9a258cfa65fa9165
fix(ext/node): `assert` compatibility (#31821)
[ { "path": "ext/node/lib.rs", "patch": "@@ -479,13 +479,10 @@ deno_core::extension!(deno_node,\n \"_process/streams.mjs\",\n \"_readline.mjs\",\n \"_util/_util_callbackify.js\",\n- \"_util/asserts.ts\",\n \"_util/async.ts\",\n \"_util/os.ts\",\n- \"_util/std_asserts.ts\",\n \"_u...
2026-01-17T09:33:31
huggingface/transformers
c620c38bb04c80bb12350bec739c298918b5675d
0798797ec9733aa8cc45f7663df90806ab51a35f
[Qwen3VLMoe] Fixed: Expected self.dtype to be equal to src.dtype - routing_weights casting (#41420) * Fixed Expected self.dtype to be equal to src.dtype on eval * Fixed Expected self.dtype to be equal to src.dtype on eval * Fixed Expected self.dtype to be equal to src.dtype on eval * generated modeling_qwen3_vl_moe...
[ { "path": "src/transformers/models/ernie4_5_moe/modeling_ernie4_5_moe.py", "patch": "@@ -307,7 +307,7 @@ def forward(\n routing_weights = routing_weights / torch.clamp(\n routing_weights.sum(dim=-1, keepdim=True), min=self.norm_min\n )\n- routing_weights = rout...
2025-10-14T11:14:49
ollama/ollama
58ffa03d8bdde2869bdd0ca892778d2130802fd4
637f8bc6a56ac83140d9f46a00d143ef95f388cc
fix impossible condition
[ { "path": "cmd/cmd.go", "patch": "@@ -114,10 +114,6 @@ func RunHandler(cmd *cobra.Command, args []string) error {\n \t}\n \n \tmp := server.ParseModelPath(args[0])\n-\tif err != nil {\n-\t\treturn err\n-\t}\n-\n \tif mp.ProtocolScheme == \"http\" && !insecure {\n \t\treturn fmt.Errorf(\"insecure protocol ht...
2023-09-19T22:23:33
ggml-org/llama.cpp
f6dcda390004b627ef30af378d0c01ad2519289e
606a73f53175077429484b23dcf799f69a31d0bd
server : context checkpointing for hybrid and recurrent models (#16382) * initial commit for branch 3 * generalize `swa_checkpoint` to `ctx_checkpoint` this extends `llama-server`'s SWA checkpointing logic to include hybrid/recurrent models such as Jamba, Granite * oops * disable debug prints * keep backwards com...
[ { "path": "common/arg.cpp", "patch": "@@ -1932,13 +1932,13 @@ common_params_context common_params_parser_init(common_params & params, llama_ex\n }\n ).set_env(\"LLAMA_ARG_SWA_FULL\"));\n add_opt(common_arg(\n- {\"--swa-checkpoints\"}, \"N\",\n- string_format(\"max number of SWA...
2025-10-03T18:34:51
vuejs/vue
bb426258a1e5892e49ed20076c972cb35164b85c
3549e73c032143854bd65c8fdca6a6299ed89e39
skip multiple select tests on Android 4.4 by detecting its bug
[ { "path": "build/karma.sauce.config.js", "patch": "@@ -64,10 +64,8 @@ var batches = [\n },\n sl_android_4_4: {\n base: 'SauceLabs',\n- browserName: 'Browser',\n- platform: 'Android',\n- version: '4.4',\n- device: 'Android Emulator'\n+ browserName: 'android',\n+ ve...
2017-01-13T00:36:07
huggingface/transformers
0798797ec9733aa8cc45f7663df90806ab51a35f
0566b6f5bdf474d51a34947f0957be84c42207bf
Fix an import error with PreTrainModel (#41571)
[ { "path": "src/transformers/quantizers/base.py", "patch": "@@ -32,6 +32,9 @@\n \n \n def _assign_original_dtype(module, original_dtype):\n+ # not very nice in a recursive function but it avoids a circular import\n+ from ..modeling_utils import PreTrainedModel\n+\n for child in module.children():\n...
2025-10-14T11:13:37
denoland/deno
28184e1c5bec78836accdd6e9a258cfa65fa9165
6ea647992fbb5da7971c0e62860bee9e008b57ae
fix(check): improve `@types/` package resolution for the global resolver (#31868)
[ { "path": "cli/tools/publish/unfurl.rs", "patch": "@@ -718,9 +718,7 @@ impl<TSys: SpecifierUnfurlerSys> SpecifierUnfurler<TSys> {\n match parts.first() {\n Some(DynamicTemplatePart::String { value: specifier }) => {\n // relative doesn't need to be modified\n- let is...
2026-01-16T21:25:06
ggml-org/llama.cpp
606a73f53175077429484b23dcf799f69a31d0bd
946f71ed9ade07e319859b5ce656144140e066fb
metal : fix loop bound in ggml_mem_ranges (#16412)
[ { "path": "ggml/src/ggml-metal/ggml-metal-common.cpp", "patch": "@@ -112,7 +112,7 @@ static bool ggml_mem_ranges_add_dst(ggml_mem_ranges_t mrs, const ggml_tensor * t\n }\n \n bool ggml_mem_ranges_add(ggml_mem_ranges_t mrs, const ggml_tensor * tensor) {\n- for (int i = 0; i < GGML_MAX_DIMS; i++) {\n+ f...
2025-10-03T16:18:56
vuejs/vue
3549e73c032143854bd65c8fdca6a6299ed89e39
79c1c7f0fe91981ae3c66556aa1990b48adb3310
Updated Android 4 SauceLabs configuration (#4672) * update sauce config for android SauceLabs has sunset all versions of Android 4 prior to 4.4, as they are modernizing their infrastructure: https://wiki.saucelabs.com/pages/viewpage.action?pageId=67012495 This means the previous Selendroid configuration must...
[ { "path": "build/karma.sauce.config.js", "patch": "@@ -62,10 +62,12 @@ var batches = [\n browserName: 'iphone',\n version: '9.3'\n },\n- sl_android_4_2: {\n+ sl_android_4_4: {\n base: 'SauceLabs',\n- browserName: 'android',\n- version: '4.2'\n+ browserName: 'Browse...
2017-01-12T23:45:43
huggingface/transformers
0566b6f5bdf474d51a34947f0957be84c42207bf
b3e3c3dc93f29770a768d6943c9fb9d377e5edce
Patch MistralCommonTokenizer (#41439) * Fix token_to_id and add add_generation_prompt * Fix spm download * Refactor spm * Try another possibly non-gated spm * Improve get_vocab * lint * Improve get_vocab * Add warn to piece_to_id * Improve from_pretrained raise and revert model spm * Revert fast
[ { "path": "src/transformers/tokenization_mistral_common.py", "patch": "@@ -115,11 +115,6 @@\n of returning overflowing tokens.\n return_special_tokens_mask (`bool`, *optional*, defaults to `False`):\n Whether or not to return special tokens mask information.\n- ...
2025-10-14T11:13:19
ollama/ollama
a5520bfb4270f9ae8b99ef227d475544b1407436
2627c464ba1e9b032a37babaa6400d2308f562a7
fix build
[ { "path": "cmd/cmd.go", "patch": "@@ -14,7 +14,6 @@ import (\n \t\"net/http\"\n \t\"os\"\n \t\"os/exec\"\n-\t\"path\"\n \t\"path/filepath\"\n \t\"runtime\"\n \t\"strings\"", "additions": 0, "deletions": 1, "language": "Go" }, { "path": "server/images.go", "patch": "@@ -14,7 +14,6 @@ ...
2023-09-19T17:42:20
denoland/deno
6ea647992fbb5da7971c0e62860bee9e008b57ae
df289e57f7f3d03513cb781dc2ee23c2b44d8c19
fix(ext/node): fix deadline resolution for queued timers (#31860) Fixes `test-timers-immediate-queue.js`
[ { "path": "Cargo.lock", "patch": "@@ -1884,9 +1884,9 @@ dependencies = [\n \n [[package]]\n name = \"deno_core\"\n-version = \"0.377.0\"\n+version = \"0.380.0\"\n source = \"registry+https://github.com/rust-lang/crates.io-index\"\n-checksum = \"38828fa5f4f80e9c098ae38c7a0498454cd2585e2621ba2b78a9f7bd4809946...
2026-01-16T18:40:14
ggml-org/llama.cpp
946f71ed9ade07e319859b5ce656144140e066fb
638d330246954e88dffc22ce01fec15e6894e544
llama : fix shapes for bert/mpt q/k norm (#16409)
[ { "path": "src/llama-model.cpp", "patch": "@@ -7843,6 +7843,8 @@ struct llm_build_bert : public llm_graph_context {\n }\n \n if (model.layers[il].attn_q_norm) {\n+ Qcur = ggml_reshape_2d(ctx0, Qcur, n_embd_head*n_head, n_tokens);\n+\n Qc...
2025-10-03T12:40:25