Datasets:
helloadhavan
/
github_issues

Dataset card Data Studio Files
xet
 Community
1
repo
stringclasses
15 values
fix_commit
stringlengths
40
40
buggy_commit
stringlengths
40
40
message
stringlengths
3
64.3k
files
listlengths
1
300
timestamp
timestamp[s]date
2013-03-13 20:45:00
2026-04-11 07:48:46
ggml-org/llama.cpp
4d3d455d3c8719eb8f206de328d1b9ba191efd7c
c9b1c06467aca8d30fafcab51292bcb285df5b0d
sync : whisper.cpp (ggml/1359) * ggml : Fix MKL detection by quoting BLAS_INCLUDE_DIRS (whisper/3426) * sync : whisper.cpp
[ { "path": "ggml/src/ggml-blas/CMakeLists.txt", "patch": "@@ -74,7 +74,7 @@ if (BLAS_FOUND)\n \n target_compile_options(ggml-blas PRIVATE ${BLAS_LINKER_FLAGS})\n \n- if (${BLAS_INCLUDE_DIRS} MATCHES \"mkl\" AND (${GGML_BLAS_VENDOR} MATCHES \"Generic\" OR ${GGML_BLAS_VENDOR} MATCHES \"Intel\"))\n+ i...
2025-09-29T13:49:11
vuejs/vue
84b7b82031ac70467cedcec801fde55a605e4321
b3ebfef91d72b301b85bce07c4ffad4f59bc73e2
fix SSR tests
[ { "path": "src/platforms/web/server/modules/dom-props.js", "patch": "@@ -20,14 +20,17 @@ export default function renderDOMProps (node: VNodeWithData): string {\n return res\n }\n \n+ const attrs = node.data.attrs\n for (const key in props) {\n if (key === 'innerHTML') {\n setText(node, pr...
2016-12-23T04:03:05
denoland/deno
8d47d7e24402c2ba1f50696e1293f8c7c056984c
7558cafe636c14d31adc5d3de2ecae4bdc6577ae
fix(ext/node): fix use-after-free in StatementSync JS iterator (#31770) Keep a reference of StatementSync alive to prevent GC from destroying statement handle while iterator is being used. Fixes https://github.com/denoland/deno/issues/31744
[ { "path": "ext/node/ops/sqlite/statement.rs", "patch": "@@ -692,6 +692,7 @@ impl StatementSync {\n RETURN = \"return\",\n DONE = \"done\",\n VALUE = \"value\",\n+ __STATEMENT_REF = \"__statement_ref\",\n }\n \n self.reset()?;\n@@ -797,8 +798,19 @@ impl StatementSync {\n le...
2026-01-08T10:29:58
ggml-org/llama.cpp
c9b1c06467aca8d30fafcab51292bcb285df5b0d
b6ae75afb49b23db3dfbc2b01e8aabfb047e6880
ggml : remove -dev suffix from release version (ggml/1355) This commit removes the `-dev` suffix from the version string in CMakeLists.txt and the release script. The version will now be just be formatted as `MAJOR.MINOR.PATCH`.
[ { "path": "ggml/CMakeLists.txt", "patch": "@@ -5,7 +5,6 @@ project(\"ggml\" C CXX ASM)\n set(GGML_VERSION_MAJOR 0)\n set(GGML_VERSION_MINOR 9)\n set(GGML_VERSION_PATCH 3)\n-set(GGML_VERSION_DEV \"\") # \"-dev\" for development, \"\" for releases\n set(GGML_VERSION_BASE \"${GGML_VERSION_MAJOR}.${GGML_VERSIO...
2025-09-26T15:34:42
huggingface/transformers
823fab4860ec7a5c71d8a21f834104c6deedfaa4
42d4e13a0bcaf7e7828f7b79ea2997be46a3dad8
Fix bnb fsdp loading for pre-quantized checkpoint (#41415) * fix * fix * get_param_name * fix device name
[ { "path": "src/transformers/modeling_utils.py", "patch": "@@ -761,21 +761,17 @@ def _load_state_dict_into_meta_model(\n # and then cast it to CPU to avoid excessive memory usage on each GPU\n # in comparison to the sharded model across GPUs.\n if is_fsdp_enabl...
2025-10-09T16:05:35
vuejs/vue
b3ebfef91d72b301b85bce07c4ffad4f59bc73e2
71cc0a575b88346cb10688c04140b335f8a01ba3
also bind static special attrs as props (fix #4530)
[ { "path": "src/compiler/parser/index.js", "patch": "@@ -459,6 +459,15 @@ function processAttrs (el) {\n }\n }\n addAttr(el, name, JSON.stringify(value))\n+ // #4530 also bind special attributes as props even if they are static\n+ // so that patches between dynamic/static are co...
2016-12-23T03:32:57
denoland/deno
7558cafe636c14d31adc5d3de2ecae4bdc6577ae
6dea9ab9dcde9742374defb182a6ff664a963a0f
fix: update config schema for read permission to include ignore property (#31785)
[ { "path": "cli/schemas/config-file.v1.json", "patch": "@@ -66,7 +66,7 @@\n \"type\": \"boolean\",\n \"description\": \"Allow all permissions for the program to run unrestricted.\"\n },\n- \"read\": { \"$ref\": \"#/$defs/allowDenyPermissionConfigValue\" },\n+ \"read\...
2026-01-07T23:02:10
ggml-org/llama.cpp
02463ab27b1379ff5e3d936ce8b3bfd356872ea6
adc76347d73b3d915e946efa5de8d0ad9f3904c2
ggml-backend : add root cause in error message if loading backend library fails (#16172) This PR adds additional information to an error message when loading backend library via ld_load_library() fails. This helps spotting why backend library did not load (missing library, missing dependency or unresolved symbol etc.)...
[ { "path": "ggml/src/ggml-backend-reg.cpp", "patch": "@@ -135,6 +135,10 @@ static void * dl_get_sym(dl_handle * handle, const char * name) {\n return p;\n }\n \n+static const char * dl_error() {\n+ return \"\";\n+}\n+\n #else\n \n using dl_handle = void;\n@@ -155,6 +159,11 @@ static void * dl_get_sym(...
2025-09-29T11:17:09
huggingface/transformers
42d4e13a0bcaf7e7828f7b79ea2997be46a3dad8
0eae41ad36c2a9823bf3906eb9e332f125583998
RT-Detr correct 2d positional embeddings for non-square images (#41380) * Correct 2d positional embeddings for non-square images * Simplify bug fix propagate changes to other models --------- Co-authored-by: Konstantinos Pitas <kostasp210@gmail.com> Co-authored-by: Yoni Gozlan <74535834+yonigozlan@users.noreply.git...
[ { "path": "src/transformers/models/d_fine/modeling_d_fine.py", "patch": "@@ -2064,7 +2064,7 @@ def build_2d_sincos_position_embedding(\n ):\n grid_w = torch.arange(torch_int(width), device=device).to(dtype)\n grid_h = torch.arange(torch_int(height), device=device).to(dtype)\n- gri...
2025-10-09T15:58:22
ollama/ollama
2bc06565c7ca7f784fefc0d2f037269b8af2a33d
d1c2558f7e97b44533f474f19e8d27a3a170adcf
fix empty response
[ { "path": "llm/ggml_llama.go", "patch": "@@ -497,8 +497,10 @@ func (llm *llama) Predict(ctx context.Context, prevContext []int, prompt string,\n \t\t\t\t\treturn fmt.Errorf(\"error unmarshaling llm prediction response: %v\", err)\n \t\t\t\t}\n \n-\t\t\t\tfn(api.GenerateResponse{Response: p.Content})\n-\t\t\...
2023-09-05T22:03:24
vuejs/vue
0eb8cdc7f7cff484ebe0b12a781a4ab020f82b4a
bc140de48b986b5079bb69bef276420749b93119
support v-bind object on scoped slots (fix #4529)
[ { "path": "src/compiler/codegen/index.js", "patch": "@@ -343,13 +343,19 @@ function genText (text: ASTText | ASTExpression): string {\n function genSlot (el: ASTElement): string {\n const slotName = el.slotName || '\"default\"'\n const children = genChildren(el)\n- return `_t(${slotName}${\n- childr...
2016-12-22T04:27:21
ggml-org/llama.cpp
3a2bdcda0b31e51db954551e0cd49424133a84f3
66bb7985c3fcefda7a74aae9f8321f70eb1646c4
Improve Mobile UI for dialogs and action dropdowns (#16222) * fix: Always show conversation item actions * feat: Improve Alert Dialog and Dialog mobile UI * feat: Add settings reset to default confirmation * fix: Close Edit dialog on save * chore: update webui build output * webui: implement proper z-index system...
[ { "path": "tools/server/webui/src/app.css", "patch": "@@ -39,6 +39,7 @@\n \t--sidebar-ring: oklch(0.708 0 0);\n \t--code-background: oklch(0.225 0 0);\n \t--code-foreground: oklch(0.875 0 0);\n+\t--layer-popover: 1000000;\n }\n \n .dark {", "additions": 1, "deletions": 0, "language": "CSS" }, ...
2025-09-29T08:37:20
denoland/deno
6dea9ab9dcde9742374defb182a6ff664a963a0f
39d71bef19e7d29d071d272d483c4e3e6c4ca10c
fix(task): preserve signal-based exit codes for child processes (#31789)
[ { "path": "Cargo.lock", "patch": "@@ -2918,17 +2918,16 @@ dependencies = [\n \n [[package]]\n name = \"deno_task_shell\"\n-version = \"0.26.0\"\n+version = \"0.26.2\"\n source = \"registry+https://github.com/rust-lang/crates.io-index\"\n-checksum = \"8ec9c589f733da0f5dfdf475d44cd01e2087425839a55162538d9898e...
2026-01-07T21:35:06
huggingface/transformers
0eae41ad36c2a9823bf3906eb9e332f125583998
589fc29c9da0b98b8a9cdf3b0d4cbe949886c274
Add Code World Model (CWM) (#41199) * [wip][cwm] Code World Model stubs and setup in HF Transformers * [wip] Get other things working * [wip] Working * Tokenizer pad * fix: cwm window attn * temp remove test * temp remove test * Fixes * Temporarily add auto config remapping option until VLLM 0.11 is out * Fix...
[ { "path": "docs/source/en/_toctree.yml", "patch": "@@ -1026,6 +1026,8 @@\n title: CLIPSeg\n - local: model_doc/clvp\n title: CLVP\n+ - local: model_doc/cwm\n+ title: Code World Model (CWM)\n - local: model_doc/cohere2_vision\n title: Cohere2Vision\n - lo...
2025-10-09T15:57:45
vuejs/vue
36193183e124f0f1a054170d2cefd335d45135b8
92ad0bd378d38f84691120b1caa654ad195ea724
avoid checked state being overwritten before change event trigger (fix #4521)
[ { "path": "src/platforms/web/runtime/modules/dom-props.js", "patch": "@@ -2,9 +2,6 @@\n \n import { extend, toNumber } from 'shared/util'\n \n-// check platforms/web/util/attrs.js acceptValue\n-declare type acceptValueElm = HTMLInputElement | HTMLSelectElement | HTMLOptionElement\n-\n function updateDOMProp...
2016-12-22T03:35:21
ggml-org/llama.cpp
66bb7985c3fcefda7a74aae9f8321f70eb1646c4
2f61c0f5bf8a620ca4c3872408803ab38cfb9613
fix: preserved zero values in chat settings inputs and textareas by switching to nullish coalescing for field values and default placeholders (#16312)
[ { "path": "tools/server/webui/src/lib/components/app/chat/ChatSettings/ChatSettingsFields.svelte", "patch": "@@ -30,9 +30,9 @@\n \n \t\t\t<Input\n \t\t\t\tid={field.key}\n-\t\t\t\tvalue={String(localConfig[field.key] || '')}\n+\t\t\t\tvalue={String(localConfig[field.key] ?? '')}\n \t\t\t\tonchange={(e) => o...
2025-09-29T07:08:41
ollama/ollama
06ef90c051d0637bd46d9339d6829fe4c7641cd6
e9f6df7dca85abfe473d5b74c25cb45286746483
fix parameter inheritence parameters are not inherited because they are processed differently from other layer. fix this by explicitly merging the inherited params into the new params. parameter values defined in the new modelfile will override those defined in the inherited modelfile. array lists are replaced instead...
[ { "path": "server/images.go", "patch": "@@ -276,6 +276,7 @@ func CreateModel(ctx context.Context, name string, path string, fn func(resp api\n \n \tvar layers []*LayerReader\n \tparams := make(map[string][]string)\n+\tvar sourceParams map[string]any\n \tembed := EmbeddingParams{fn: fn}\n \tfor _, c := range...
2023-09-05T18:05:03
huggingface/transformers
589fc29c9da0b98b8a9cdf3b0d4cbe949886c274
26b5b52676ec22e1cf94bc27ed1a6c542e60159d
enhance patched_tearDown to support python 3.11+ (#41429) * enhance to support python 3.11+ Signed-off-by: Yao, Matrix <matrix.yao@intel.com> * fix style Signed-off-by: Yao, Matrix <matrix.yao@intel.com> --------- Signed-off-by: Yao, Matrix <matrix.yao@intel.com>
[ { "path": "src/transformers/testing_utils.py", "patch": "@@ -3533,8 +3533,15 @@ def _patched_tearDown(self, *args, **kwargs):\n # We still record those failures not handled by the patched methods, and add custom messages along with the usual\n # pytest failure report.\n regular_failures_info = [...
2025-10-09T15:49:29
denoland/deno
39d71bef19e7d29d071d272d483c4e3e6c4ca10c
162e9e483f2fe4e26ff3ab7ec8cb6ac0c773a061
fix(npm): deduplication pass for initially partially resolved graphs (#31787) * https://github.com/denoland/deno_npm/pull/130
[ { "path": "Cargo.lock", "patch": "@@ -2530,9 +2530,9 @@ dependencies = [\n \n [[package]]\n name = \"deno_npm\"\n-version = \"0.42.2\"\n+version = \"0.42.3\"\n source = \"registry+https://github.com/rust-lang/crates.io-index\"\n-checksum = \"7daa908c280b25d68ea59765144df488714716cc14a2dcafaa1ec3906d4abbbe\"...
2026-01-07T21:29:21
vuejs/vue
92ad0bd378d38f84691120b1caa654ad195ea724
dacd0cf58241c66c1946cef21475903aa18cad0e
ensure leave transitions and enter transitions are triggered in the same frame (fix #4510)
[ { "path": "examples/move-animations/index.html", "patch": "@@ -15,17 +15,19 @@\n border: 1px solid #666;\n box-sizing: border-box;\n }\n+ /* 1. define transition property, duration and easing */\n .fade-move, .fade-enter-active, .fade-leave-active {\n transition: all...
2016-12-22T03:03:50
ggml-org/llama.cpp
2f61c0f5bf8a620ca4c3872408803ab38cfb9613
3ffd0fae473c954bb3e67526b31262048fb508d4
llama-cli: prevent spurious assistant token (#16202) * tools/main: llama-cli: prevent spurious assistant token (#13402) During prompt ingestion, prompt tokens are accepted into the sampler history (for repetition penalties). The conversation-mode path then appended `common_sampler_last(smpl)` to `assistant_ss` before...
[ { "path": "tools/main/main.cpp", "patch": "@@ -707,6 +707,10 @@ int main(int argc, char ** argv) {\n \n embd.push_back(id);\n \n+ if (params.conversation_mode && !waiting_for_first_input && !llama_vocab_is_eog(vocab, id)) {\n+ assistant_ss << common_token_to_piece(ctx, ...
2025-09-29T07:03:12
huggingface/transformers
26b5b52676ec22e1cf94bc27ed1a6c542e60159d
34b861abd11074fd32362b9a25c1cc582fa0b941
[Fix] Fix test file error (#40973) Fix test file error
[ { "path": "tests/trainer/test_trainer.py", "patch": "@@ -5907,7 +5907,7 @@ def hp_name(trial):\n )\n optim_test_params.append(\n (\n- TrainingArguments(optim=OptimizerNames.ADAMW_TORCH_8BIT, output_dir=\"None\"),\n+ OptimizerNames.ADAMW_TORCH_8BIT,\n...
2025-10-09T15:30:53
denoland/deno
162e9e483f2fe4e26ff3ab7ec8cb6ac0c773a061
014ba7950152d64c151064f5533e75c1a182622c
fix(webgpu): correctly expose `GPUCompilationInfo` & `GPUCompilationMessage` (#31780) Closes #28649
[ { "path": "cli/tsc/dts/lib.deno_webgpu.d.ts", "patch": "@@ -747,15 +747,17 @@ interface GPUPipelineLayoutDescriptor extends GPUObjectDescriptorBase {\n type GPUCompilationMessageType = \"error\" | \"warning\" | \"info\";\n \n /** @category GPU */\n-interface GPUCompilationMessage {\n+declare class GPUCompil...
2026-01-07T20:54:00
vuejs/vue
451ea37c7b75e18a663f978b3d30fec214eb68b6
56bfa1d291e7cf8c3e70386ce936ab1965b545ed
don`t cast empty string to Boolean if prop types include String and Boolean (fix #4538) (#4539)
[ { "path": "src/core/util/props.js", "patch": "@@ -21,10 +21,10 @@ export function validateProp (\n const absent = !hasOwn(propsData, key)\n let value = propsData[key]\n // handle boolean props\n- if (isBooleanType(prop.type)) {\n+ if (isType(Boolean, prop.type)) {\n if (absent && !hasOwn(prop, '...
2016-12-21T21:41:52
denoland/deno
014ba7950152d64c151064f5533e75c1a182622c
dfb4e701bbe7e313c756628553c692f214cff41e
fix(cli/test): close Node HTTP connections before sanitizer check (#31782) Closes https://github.com/denoland/deno/issues/31777 Fixes false positive "leaking resources" errors when running tests that use Node.js HTTP/HTTPS client with connection pooling. - Add `closeIdleConnections()` to close HTTP Agent pooled conn...
[ { "path": "cli/tools/test/mod.rs", "patch": "@@ -1126,6 +1126,14 @@ async fn run_tests_for_worker_inner(\n continue;\n }\n \n+ // Close idle Node.js HTTP Agent connections to prevent cross-test\n+ // pollution and false positive resource leak detection from pooled\n+ // keepAlive connecti...
2026-01-07T20:45:11
vuejs/vue
be9210fc78339ba1fb543160a4b18e8fc444ca68
a977642fac385b2d16455a5eaebf70f2fc949377
update dom prop for option.value (fix #4494) (#4505) * update dom prop for option.value * refactor value update logic
[ { "path": "src/platforms/web/runtime/modules/dom-props.js", "patch": "@@ -2,6 +2,9 @@\n \n import { extend, toNumber } from 'shared/util'\n \n+// check platforms/web/util/attrs.js acceptValue\n+declare type acceptValueElm = HTMLInputElement | HTMLSelectElement | HTMLOptionElement\n+\n function updateDOMProp...
2016-12-21T21:19:04
ollama/ollama
681f3c4c42a645fb3e9cbc4311ae6662d838a684
5d3f314b0bdcc7a599f65e947bee65e3cc4c73bd
fix num_keep
[ { "path": "server/routes.go", "patch": "@@ -117,12 +117,13 @@ func load(ctx context.Context, model *Model, reqOpts map[string]interface{}, ses\n \t\t\tif err != nil {\n \t\t\t\treturn err\n \t\t\t}\n+\n \t\t\ttokensNoSystem, err := llmModel.Encode(ctx, promptNoSystem)\n \t\t\tif err != nil {\n \t\t\t\tretur...
2023-09-03T21:36:14
huggingface/transformers
34b861abd11074fd32362b9a25c1cc582fa0b941
b44d91570fb4a5cf8a58150c7935a4edde82e27f
:rotating_light: [`Attention Masks`] Bidirectional masks for encoder and encoder-decoder models (#41265) * new masks * fixes * adjust comments * fix unnecessary mask creation on sdpa * simplify masks more * propogate to other models * style + repo consistency * copies * no comment * fix attempt * finally fix...
[ { "path": "src/transformers/integrations/executorch.py", "patch": "@@ -26,6 +26,7 @@\n from ..generation.configuration_utils import GenerationConfig\n from ..masking_utils import (\n ALL_MASK_ATTENTION_FUNCTIONS,\n+ _ignore_bidirectional_mask_sdpa,\n _ignore_causal_mask_sdpa,\n _is_torch_grea...
2025-10-09T14:56:11
denoland/deno
dfb4e701bbe7e313c756628553c692f214cff41e
df6daeeaa9528fbd619ec162c7e3081551e596f0
fix(tsgo): nil deref for exported symbols (#31786)
[ { "path": "cli/tsc/go/tsgo_version.rs", "patch": "@@ -23,15 +23,15 @@ impl Hashes {\n }\n }\n \n-pub const VERSION: &str = \"0.1.13\";\n+pub const VERSION: &str = \"0.1.14\";\n pub const DOWNLOAD_BASE_URL: &str =\n- \"https://github.com/denoland/typescript-go/releases/download/v0.1.13\";\n+ \"https://gi...
2026-01-07T19:50:23
ggml-org/llama.cpp
2f61c0f5bf8a620ca4c3872408803ab38cfb9613
3ffd0fae473c954bb3e67526b31262048fb508d4
llama-cli: prevent spurious assistant token (#16202) * tools/main: llama-cli: prevent spurious assistant token (#13402) During prompt ingestion, prompt tokens are accepted into the sampler history (for repetition penalties). The conversation-mode path then appended `common_sampler_last(smpl)` to `assistant_ss` before...
[ { "path": "tools/main/main.cpp", "patch": "@@ -707,6 +707,10 @@ int main(int argc, char ** argv) {\n \n embd.push_back(id);\n \n+ if (params.conversation_mode && !waiting_for_first_input && !llama_vocab_is_eog(vocab, id)) {\n+ assistant_ss << common_token_to_piece(ctx, ...
2025-09-29T07:03:12
ollama/ollama
59a705525c91bfa407c2f5fa58eac33c109bdd4a
5d3f314b0bdcc7a599f65e947bee65e3cc4c73bd
fix not forwarding last token
[ { "path": "llm/ggml_llama.go", "patch": "@@ -353,11 +353,6 @@ func (llm *llama) SetOptions(opts api.Options) {\n \tllm.Options = opts\n }\n \n-type Prediction struct {\n-\tContent string `json:\"content\"`\n-\tStop bool `json:\"stop\"`\n-}\n-\n type GenerationSettings struct {\n \tFrequencyPenalty floa...
2023-09-03T21:46:35
vuejs/vue
a977642fac385b2d16455a5eaebf70f2fc949377
5c34b1ba731ffc53aa10c95bbf5d943f0415372f
unbind v-show if no longer present during patch (fix #4484)
[ { "path": "src/core/vdom/modules/directives.js", "patch": "@@ -20,6 +20,7 @@ function updateDirectives (oldVnode: VNodeWithData, vnode: VNodeWithData) {\n \n function _update (oldVnode, vnode) {\n const isCreate = oldVnode === emptyNode\n+ const isDestroy = vnode === emptyNode\n const oldDirs = normali...
2016-12-21T21:08:34
huggingface/transformers
b44d91570fb4a5cf8a58150c7935a4edde82e27f
d99069195b11824e35fdffa0319baea266b8a885
[v5] remove load_in_4bit and load_in_8bit (#41287) * [v5] remove load_in_4bit and load_in_8bit * fix * reveert * fix --------- Co-authored-by: Mohamed Mekkouri <93391238+MekkCyber@users.noreply.github.com>
[ { "path": "docs/source/ar/llm_tutorial.md", "patch": "@@ -60,10 +60,10 @@ pip install transformers bitsandbytes>=0.39.0 -q\n أولاً، تحتاج إلى تحميل النموذج.\n \n ```py\n->>> from transformers import AutoModelForCausalLM\n+>>> from transformers import AutoModelForCausalLM, BitsAndBytesConfig\n \n >>> model =...
2025-10-09T14:34:04
denoland/deno
cfceaa82d069dd152de3b2e46fc4a9c4b9e9c10e
ad4d0649b66c3871199890192e9b85499c657662
fix: queue dynamic imports for modules with pending TLA (#31670) When a module with Top-Level Await (TLA) is dynamically imported multiple times before the TLA resolves, subsequent imports would fail with "Cannot access 'default' before initialization". This happened because V8 sets module status to `Evaluated` immed...
[ { "path": "Cargo.lock", "patch": "@@ -1883,9 +1883,9 @@ dependencies = [\n \n [[package]]\n name = \"deno_core\"\n-version = \"0.376.0\"\n+version = \"0.377.0\"\n source = \"registry+https://github.com/rust-lang/crates.io-index\"\n-checksum = \"829ee94e76838e99951cc7b3ccd31f68a220687f52229c0d6e135ff29b22f6d...
2026-01-07T15:16:30
ggml-org/llama.cpp
a4a0aa5ea2a88e4858996199fefd5439f17b481c
92cd103f627015a95f2107f1c27dc218c7b8ec64
ggml : fix dependencies for ggml_set_rows (#16318)
[ { "path": "ggml/src/ggml.c", "patch": "@@ -3687,6 +3687,7 @@ struct ggml_tensor * ggml_set_rows(\n result->op = GGML_OP_SET_ROWS;\n result->src[0] = b;\n result->src[1] = c;\n+ result->src[2] = a; // note: order is weird due to legacy reasons (https://github.com/ggml-org/llama.cpp/pull/16...
2025-09-29T05:41:28
vuejs/vue
5c34b1ba731ffc53aa10c95bbf5d943f0415372f
d212a109b0d588ea67a3707d92ee997dccd1e828
fix type definition links (#4517) * fix type definition links NOTE: Everytime, we tried not to change links. * update flow links
[ { "path": "packages/vue-template-compiler/README.md", "patch": "@@ -38,7 +38,7 @@ The optional `options` object can contain the following:\n \n - `modules`\n \n- An array of compiler modules. For details on compiler modules, refer to its [type definition](https://github.com/vuejs/vue/blob/dev/flow/compiler...
2016-12-21T03:30:54
denoland/deno
ad4d0649b66c3871199890192e9b85499c657662
5a5d63915cf8e78cd9abb757c849597010328931
chore: Enable windows-sys features in specific members (#31679) This change enables additional `windows-sys` features (beyond those declared in the workspace Cargo.toml) needed by each member. Fixes #31678
[ { "path": "ext/node/Cargo.toml", "patch": "@@ -108,6 +108,6 @@ errno = \"0.3.10\"\n nix = { workspace = true, features = [\"user\"] }\n \n [target.'cfg(windows)'.dependencies]\n-windows-sys.workspace = true\n+windows-sys = { workspace = true, features = [\"Win32_Networking_WinSock\"] }\n winapi = { workspac...
2026-01-07T15:13:58
huggingface/transformers
d99069195b11824e35fdffa0319baea266b8a885
bf38b2d11dd17c61a27ea91fac53f8ad92e3171e
Cleaning hub kernels (#41477) * disable kernel mapping * cleaning * revert * fix style
[ { "path": "src/transformers/generation/continuous_batching/continuous_api.py", "patch": "@@ -27,7 +27,7 @@\n \n from ...configuration_utils import PreTrainedConfig\n from ...generation.configuration_utils import GenerationConfig\n-from ...integrations.hub_kernels import load_and_register_kernel\n+from ...in...
2025-10-09T14:32:18
ggml-org/llama.cpp
92cd103f627015a95f2107f1c27dc218c7b8ec64
b887d2f3413ac231e3cb5925260c39902af4a70c
vulkan: Fix validation failure in quantized flash attention (#16292)
[ { "path": "ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_base.comp", "patch": "@@ -67,30 +67,48 @@ layout (binding = 5) writeonly buffer O {D_TYPE data_o[];};\n #if defined(A_TYPE_PACKED16)\n #define BINDING_IDX_K 0\n #define BINDING_IDX_V 1\n-layout (binding = 1) readonly buffer KV_PACKED16 {A_TYPE_PACKED...
2025-09-29T04:50:37
vuejs/vue
974247fd8e6b57840c9b7686964ddb38bc6affc8
34333caa8cb54a4eacc723fb46c18334ca8a7c11
Add missing string handler in v-for (#4499) Fix #4497
[ { "path": "src/core/instance/render.js", "patch": "@@ -178,7 +178,7 @@ export function renderMixin (Vue: Class<Component>) {\n render: () => VNode\n ): ?Array<VNode> {\n let ret: ?Array<VNode>, i, l, keys, key\n- if (Array.isArray(val)) {\n+ if (Array.isArray(val) || typeof val === 'string')...
2016-12-16T16:50:53
huggingface/transformers
bf38b2d11dd17c61a27ea91fac53f8ad92e3171e
72a3fc275c6161fcb98664c81aae8038a76407ce
Change RT-Detr docs to reflect fixed 640x640 input size (#41364) * Update rt_detr docs to mention 640x640 input size The authors of RT-Detr mention that the model was trained on 640x640 images and was meant to be used for inference on 640x640 images. Also, the current implementation has certain quirks that make train...
[ { "path": "docs/source/en/model_doc/rt_detr.md", "patch": "@@ -40,8 +40,7 @@ The model version was contributed by [rafaelpadilla](https://huggingface.co/rafa\n \n ## Usage tips\n \n-Initially, an image is processed using a pre-trained convolutional neural network, specifically a Resnet-D variant as referenc...
2025-10-09T14:29:16
denoland/deno
5a5d63915cf8e78cd9abb757c849597010328931
4d747eafcc1e373fe2cb9d82e0ab4dc333357b48
docs(cli): add DENO_KV_DEFAULT_PATH and DENO_KV_PATH_PREFIX to help (#31720) Closes #31669
[ { "path": "cli/args/flags.rs", "patch": "@@ -1439,6 +1439,16 @@ static ENV_VARS: &[EnvVar] = &[\n description: \"Controls whether Deno.openKv() API should use disk based or in-memory\\ndatabase.\",\n example: None,\n },\n+ EnvVar {\n+ name: \"DENO_KV_DEFAULT_PATH\",\n+ description: \"Set th...
2026-01-07T15:10:13
ggml-org/llama.cpp
b887d2f3413ac231e3cb5925260c39902af4a70c
bd0af02fc96c2057726f33c0f0daf7bb8f3e462a
ggml : fix GGML_F32_VEC_FMA argument order in ggml_vec_mad1_f32 (#16307) * fix GGML_F32_VEC_FMA argument order in ggml_vec_mad1_f32 * add test that fails on simd
[ { "path": "ggml/src/ggml-cpu/vec.h", "patch": "@@ -610,7 +610,7 @@ inline static void ggml_vec_mad1_f32(const int n, float * y, const float * x, co\n for (int i = 0; i < np; i += GGML_F32_STEP) {\n for (int j = 0; j < GGML_F32_ARR; j++) {\n ay[j] = GGML_F32_VEC_LOAD(x + i...
2025-09-28T21:15:03
vuejs/vue
bf8aac0053d049a54cd72e01a85b032a58039ac6
7f260e1185284f6b1fea35bfacbc06f705ea2f06
fix weex test case
[ { "path": "test/weex/compiler/compile.spec.js", "patch": "@@ -38,7 +38,7 @@ describe('compile basic', () => {\n <text style=\"margin-left:36px;color:#eee;\">Load more...</text>\n </refresh>\n `)\n- expect(render).toEqual(`with(this){return _c('refresh',{staticClass:[\"refresh\"],staticS...
2016-12-14T17:29:30
huggingface/transformers
72a3fc275c6161fcb98664c81aae8038a76407ce
9ef804472b25c4f69c1eb213dea6f791615538a0
Remove infer_device (#41088) * Remove infer_device Signed-off-by: Yuanyuan Chen <cyyever@outlook.com> * Fix docs using accelerator Signed-off-by: Yuanyuan Chen <cyyever@outlook.com> * Fix conflict Signed-off-by: Yuanyuan Chen <cyyever@outlook.com> --------- Signed-off-by: Yuanyuan Chen <cyyever@outlook.com>
[ { "path": "docs/source/en/cache_explanation.md", "patch": "@@ -98,9 +98,10 @@ The example below demonstrates how to create a generation loop with [`DynamicCac\n \n ```py\n import torch\n-from transformers import AutoTokenizer, AutoModelForCausalLM, DynamicCache, infer_device\n+from transformers import AutoT...
2025-10-09T14:05:39
denoland/deno
4d747eafcc1e373fe2cb9d82e0ab4dc333357b48
f47ad1f05a813a6fd8bc15dde979d23e3bdf6af8
fix(ext/node): use correct block sizes for HMAC algorithms (#31775) The HMAC implementation was hardcoded to use 64-byte blocks for most algorithms and 128-byte blocks only for sha512/sha384. This is incorrect for many algorithms like sha3-* variants and blake2 which have different block sizes. Add a lookup table wit...
[ { "path": "ext/node/polyfills/internal/crypto/hash.ts", "patch": "@@ -45,6 +45,7 @@ import {\n import LazyTransform from \"ext:deno_node/internal/streams/lazy_transform.js\";\n import {\n getDefaultEncoding,\n+ getHashBlockSize,\n toBuf,\n } from \"ext:deno_node/internal/crypto/util.ts\";\n import {\n@...
2026-01-06T19:49:57
ollama/ollama
1c8fd627ad9e96f58bad346d5f5a13750700f76a
ae950b00f19ccef2c02523cc9519de535500d9b3
windows: fix create modelfile
[ { "path": "server/images.go", "patch": "@@ -235,8 +235,8 @@ func GetModel(name string) (*Model, error) {\n \n func filenameWithPath(path, f string) (string, error) {\n \t// if filePath starts with ~/, replace it with the user's home directory.\n-\tif strings.HasPrefix(f, \"~/\") {\n-\t\tparts := strings.Spl...
2023-08-30T20:01:23
ggml-org/llama.cpp
bd0af02fc96c2057726f33c0f0daf7bb8f3e462a
d9e0e7c8194dfd7d23bf3a86608c9ece68d77c93
common : fix reasoning before forced tool call via tool_choice = required (#16264) * common : fix reasoning before forced tool call via tool_choice = required * common : improve reasoning and commentary handling when tool_choice is required (cherry picked from commit c746984956d6882c2de73d53ae2bb3bdf889e475) ------...
[ { "path": "common/chat.cpp", "patch": "@@ -1616,17 +1616,36 @@ static common_chat_params common_chat_params_init_gpt_oss(const common_chat_temp\n );\n });\n \n- auto recipient_in_role = builder.add_rule(\"recipient_in_role\",\n- \"\\\"<|start|>assistant\...
2025-09-28T18:13:50
vuejs/vue
f1c38674f28116e1e335873ecbc6716fb70448a3
6116bbf13a8d4827d28874636489b02399f53858
use simple normalization for components, fix functional component multi-root node (fix #4472)
[ { "path": "src/compiler/codegen/index.js", "patch": "@@ -3,13 +3,14 @@\n import { genHandlers } from './events'\n import { baseWarn, pluckModuleFunction } from '../helpers'\n import baseDirectives from '../directives/index'\n-import { camelize } from 'shared/util'\n+import { camelize, no } from 'shared/util...
2016-12-14T17:22:54
huggingface/transformers
add4df62ba7b68ed26e9828c02aa6f886e61caa1
3e8707266646dd66fbef65dad1ed2ae6d30cab8b
Fix tests fsdp (#41422) * Fix tests * fix ! * fix
[ { "path": "src/transformers/training_args.py", "patch": "@@ -473,7 +473,7 @@ class TrainingArguments:\n When resuming training, whether or not to skip the epochs and batches to get the data loading at the same\n stage as in the previous training. If set to `True`, the training will b...
2025-10-09T12:09:52
ollama/ollama
ae950b00f19ccef2c02523cc9519de535500d9b3
eeb40a672c80f0cc06b08faaebec67b954202e4f
windows: fix delete
[ { "path": "server/images.go", "patch": "@@ -859,38 +859,38 @@ func DeleteModel(name string) error {\n \tif err != nil {\n \t\treturn err\n \t}\n-\terr = filepath.Walk(fp, func(path string, info os.FileInfo, err error) error {\n-\t\tif err != nil {\n-\t\t\treturn err\n+\n+\twalkFunc := func(path string, info...
2023-08-30T18:31:12
denoland/deno
0490c42da4f78f2ca75bae608ffdc654b3e3095b
202cb32f8022ba924b778a0d60e658ebc529985b
fix(ext/node): handle errors in sqlite session filter callback and ad… (#31756) Closes https://github.com/denoland/deno/issues/31635 Fixes two issues in the `node:sqlite` module: 1. The filter handler in `applyChangeset` was panicking when the JavaScript filter callback threw an exception. The code used `.unwrap()` ...
[ { "path": "ext/node/ops/sqlite/database.rs", "patch": "@@ -963,6 +963,7 @@ impl DatabaseSync {\n scope: &'a mut v8::PinScope<'b, 'c>,\n confict: Option<v8::Local<'b, v8::Function>>,\n filter: Option<v8::Local<'b, v8::Function>>,\n+ has_caught: bool,\n }\n \n // Conflict handle...
2026-01-05T14:04:15
vuejs/vue
6116bbf13a8d4827d28874636489b02399f53858
0fe431b490b90a412ade6e8c4b8065b1f4189b3b
fix svg foreignObject regression (fix #4478)
[ { "path": "src/core/vdom/create-element.js", "patch": "@@ -71,7 +71,6 @@ export function _createElement (\n // unknown or unlisted namespaced elements\n // check at runtime because it may get assigned a namespace when its\n // parent normalizes children\n- ns = tag === 'foreignObject'...
2016-12-14T16:24:53
ggml-org/llama.cpp
d9e0e7c8194dfd7d23bf3a86608c9ece68d77c93
0124ac989f7e7bf08803788f66dbe4106bdcdd58
ci : fix musa docker build (#16306) Signed-off-by: Xiaodong Ye <yeahdongcn@gmail.com>
[ { "path": ".devops/musa.Dockerfile", "patch": "@@ -2,9 +2,9 @@ ARG UBUNTU_VERSION=22.04\n # This needs to generally match the container host's environment.\n ARG MUSA_VERSION=rc4.3.0\n # Target the MUSA build image\n-ARG BASE_MUSA_DEV_CONTAINER=sh-harbor.mthreads.com/haive/mthreads/musa:${MUSA_VERSION}-deve...
2025-09-28T14:38:15
huggingface/transformers
3e8707266646dd66fbef65dad1ed2ae6d30cab8b
f0544d7e7c631aa4629ed5d687bf4d13a9229569
Fix auto model configuration for encoder of perceptionlm (#41464) * fix auto model configuration for encoder of perceptionlm * delete perception_encoder auto registrations
[ { "path": "src/transformers/models/auto/configuration_auto.py", "patch": "@@ -307,7 +307,6 @@\n (\"pegasus\", \"PegasusConfig\"),\n (\"pegasus_x\", \"PegasusXConfig\"),\n (\"perceiver\", \"PerceiverConfig\"),\n- (\"perception_encoder\", \"TimmWrapperConfig\"),\n (\"per...
2025-10-09T12:08:03
denoland/deno
202cb32f8022ba924b778a0d60e658ebc529985b
f99eb8ce89ca4091ae94182b740fc6c345c070d1
fix(node/crypto): integrate async_hooks for domain error handling (#31751) Closes https://github.com/denoland/deno/issues/31636 This PR adds: - `AsyncHook` in `domain.ts` with `init`/`before`/`after`/`destroy` callbacks to associate async operations with domains - `emitInit`/`emitBefore`/`emitAfter`/`emitDestroy` cal...
[ { "path": "ext/node/polyfills/domain.ts", "patch": "@@ -5,8 +5,10 @@\n \n import { primordials } from \"ext:core/mod.js\";\n import { ERR_UNHANDLED_ERROR } from \"ext:deno_node/internal/errors.ts\";\n+import { AsyncHook } from \"ext:deno_node/internal/async_hooks.ts\";\n const {\n ArrayPrototypeIndexOf,\n...
2026-01-05T14:04:06
ollama/ollama
eeb40a672c80f0cc06b08faaebec67b954202e4f
0f541a03677863026ffa17367359a7f0bbfb6afe
fix list models for windows
[ { "path": "server/modelpath.go", "patch": "@@ -46,7 +46,7 @@ func ParseModelPath(name string) ModelPath {\n \t\tname = after\n \t}\n \n-\tparts := strings.Split(name, \"/\")\n+\tparts := strings.Split(name, string(os.PathSeparator))\n \tswitch len(parts) {\n \tcase 3:\n \t\tmp.Registry = parts[0]", "add...
2023-08-30T18:14:12
ggml-org/llama.cpp
2811c65286ae954bec87049f75b86dc022006dcc
d8359f5fde480da030bf75c7711573c7c4d993ba
Fixed a few typos in the README of the LLaMA.cpp HTTP Server [no ci] (#16297)
[ { "path": "tools/server/README.md", "patch": "@@ -391,7 +391,7 @@ node index.js\n \n ## API Endpoints\n \n-### GET `/health`: Returns heath check result\n+### GET `/health`: Returns health check result\n \n This endpoint is public (no API key check).\n \n@@ -846,7 +846,7 @@ To use this endpoint with POST me...
2025-09-28T11:04:46
vuejs/vue
4cca50725a5a204f7bb5cd4c7591dccf25894414
6918436bf85ec49823d2517f933fea96391f9200
fix v-for on v-else branch regression (fix #4464)
[ { "path": "src/compiler/codegen/index.js", "patch": "@@ -298,14 +298,18 @@ function genChildren (el: ASTElement, checkSkip?: boolean): string | void {\n function canSkipNormalization (children): boolean {\n for (let i = 0; i < children.length; i++) {\n const el: any = children[i]\n- if (el.for || e...
2016-12-13T17:02:37
huggingface/transformers
f0544d7e7c631aa4629ed5d687bf4d13a9229569
d1c6310d6a02481d48d81607cba7840be04580d1
Remove KERAS_NLP_IMPORT_ERROR (#41468) Remove unused variables of error messages Signed-off-by: Yuanyuan Chen <cyyever@outlook.com>
[ { "path": "src/transformers/utils/import_utils.py", "patch": "@@ -1511,12 +1511,6 @@ def is_training_run_on_sagemaker() -> bool:\n `pip install scipy`. Please note that you may need to restart your runtime after installation.\n \"\"\"\n \n-# docstyle-ignore\n-KERAS_NLP_IMPORT_ERROR = \"\"\"\n-{0} requires t...
2025-10-09T11:58:30
denoland/deno
090c7f3b276ce2a0598a44e67f8f51d6356d5c82
ded26474c0112b4b6817a1e5f8874ac97e088d8e
fix: deno cross compile for intel mac target (#31714) Do not attempt to codesign the final deno compile binary on non-macOS hosts. This restores earlier behavior of deno compile when binary is built for the intel mac target Fixes https://github.com/denoland/deno/issues/31698 Fixes https://github.com/denoland/deno/iss...
[ { "path": "Cargo.lock", "patch": "@@ -5812,9 +5812,9 @@ dependencies = [\n \n [[package]]\n name = \"libsui\"\n-version = \"0.12.5\"\n+version = \"0.12.6\"\n source = \"registry+https://github.com/rust-lang/crates.io-index\"\n-checksum = \"2d68355a1ba8740fde49a5dbd4ba3e0f3f36dbd122a39f8a3e362e43107226ff\"\n...
2026-01-05T13:09:32
ollama/ollama
1363f537ce0331ab6c09238795960a01c8560d36
bc3e21fdc67f7d933bf4930b1ce8b38a0a585b0b
fix spelling errors in prompt
[ { "path": "examples/dockerit/Modelfile", "patch": "@@ -1,6 +1,6 @@\n FROM llama2\n SYSTEM \"\"\"\n-You are an experience Devops engineer focused on docker. When given specifications for a particular need or application you know the best way to host that within a docker container. For instance if someone tel...
2023-08-31T09:02:46
ggml-org/llama.cpp
d8359f5fde480da030bf75c7711573c7c4d993ba
6a2c6145a0b91b40eb3c3dba7b20ccc4b270490f
vulkan: 64-bit im2col (#16135) * vulkan: 64-bit im2col Add variants of the im2col shaders that use buffer_device_address/buffer_reference, and use 64-bit address calculations. This is needed for large convolutions used in stable-diffusion.cpp. * fix validation error for large im2col
[ { "path": "ggml/src/ggml-vulkan/ggml-vulkan.cpp", "patch": "@@ -408,6 +408,8 @@ struct vk_device_struct {\n bool subgroup_ballot;\n bool subgroup_clustered;\n bool multi_add;\n+ bool shader_int64;\n+ bool buffer_device_address;\n \n bool add_rms_fusion;\n uint32_t partials_binding_...
2025-09-28T06:38:37
vuejs/vue
6918436bf85ec49823d2517f933fea96391f9200
5300e717fa52480bf301453beabbd863eeff0d8c
Update normalize children (fix 4466) (#4468) * omit boolean node * add test case * update boolean type * update test case * update test case
[ { "path": "src/core/vdom/helpers/normalize-children.js", "patch": "@@ -16,7 +16,7 @@ function normalizeArrayChildren (children: any, nestedIndex?: string): Array<VNo\n let i, c, last\n for (i = 0; i < children.length; i++) {\n c = children[i]\n- if (c == null) continue\n+ if (c == null || type...
2016-12-13T16:52:25
denoland/deno
7222e85d435b977de1ab810db067b86f29e6444f
e5b1487f7f6dc18b8ce8618ba3f8455f8379ca8b
fix(ext/websocket): prevent sending multiple close frames (#31747) Fixes https://github.com/denoland/deno/issues/21642
[ { "path": "Cargo.lock", "patch": "@@ -3975,7 +3975,7 @@ source = \"registry+https://github.com/rust-lang/crates.io-index\"\n checksum = \"27cea6e7f512d43b098939ff4d5a5d6fe3db07971e1d05176fe26c642d33f5b8\"\n dependencies = [\n \"getrandom 0.3.3\",\n- \"rand 0.9.1\",\n+ \"rand 0.9.2\",\n \"siphasher 1.0.1\"...
2026-01-02T18:09:39
ollama/ollama
f4432e1dbac6ca8af83979bd324cc0483fc8db7a
982c535428efc1daa0d0ed82f004f0982932a99e
treat stop as stop sequences, not exact tokens (#442) The `stop` option to the generate API is a list of sequences that should cause generation to stop. Although these are commonly called "stop tokens", they do not necessarily correspond to LLM tokens (per the LLM's tokenizer). For example, if the caller sends a gener...
[ { "path": "docs/modelfile.md", "patch": "@@ -123,7 +123,7 @@ PARAMETER <parameter> <parametervalue>\n | repeat_last_n | Sets how far back for the model to look back to prevent repetition. (Default: 64, 0 = disabled, -1 = num_ctx) ...
2023-08-30T15:53:42
ggml-org/llama.cpp
1384abf8b8d5894d32fada453ccf4d196ffba7de
e6d65fb02d553bd79cad94e517cdca18b687788d
vulkan: handle mat_mul with A matrix > 4GB (#16176) * vulkan: handle mat_mul with A matrix > 4GB This change splits mat_mul operations with huge A matrix into chunks in the M dimension. This works well for stable-diffusion use cases where the im2col matrix has very large M. Fix the order of setting the stride in mul...
[ { "path": "ggml/src/ggml-vulkan/ggml-vulkan.cpp", "patch": "@@ -5661,8 +5661,12 @@ static void ggml_vk_buffer_memset(vk_buffer& dst, size_t offset, uint32_t c, siz\n ggml_vk_queue_command_pools_cleanup(dst->device);\n }\n \n-static uint32_t ggml_vk_guess_split_k(ggml_backend_vk_context * ctx, uint32_t m...
2025-09-28T01:36:34
vuejs/vue
5300e717fa52480bf301453beabbd863eeff0d8c
68abdc3c3c86e01089c32b434e662dc15a3a9eda
Fix unkpg.com -> unpkg.com (#4460)
[ { "path": "dist/README.md", "patch": "@@ -7,7 +7,7 @@\n The full (compiler-included) browser build. This is the build you can just include with a script tag:\n \n ```\n- <script src=\"https://unkpg.com/vue/dist/vue.js\"><script>\n+ <script src=\"https://unpkg.com/vue/dist/vue.js\"></script>\n ```\n ...
2016-12-13T05:00:20
huggingface/transformers
d1c6310d6a02481d48d81607cba7840be04580d1
927aa8bef2f29296a34840b3562f9c03cc45ef81
🚨 [v5] Rendundant code in nested configs (#41314) * batch update models * delete even more * fix modular super init location * fix * fix copies * fix again, these have force-set values in configs * fix copies
[ { "path": "docs/source/en/model_doc/align.md", "patch": "@@ -154,7 +154,6 @@ for label, score in zip(candidate_labels, probs):\n ## AlignConfig\n \n [[autodoc]] AlignConfig\n- - from_text_vision_configs\n \n ## AlignTextConfig\n ", "additions": 0, "deletions": 1, "language": "Markdown" }, ...
2025-10-09T11:47:44
ggml-org/llama.cpp
0499b29c6f64c705faaf5860dc4600fca23671f4
234e2ff8ed09716fb553437596779399bee31b11
vulkan: throw system error instead of SIGABRT during init on older devices (#16156) * Throw system error on old Vulkan driver rather than SIGABRT * Optionally handle any potential error in vulkan init
[ { "path": "ggml/src/ggml-vulkan/ggml-vulkan.cpp", "patch": "@@ -4521,7 +4521,7 @@ static void ggml_vk_instance_init() {\n \n if (api_version < VK_API_VERSION_1_2) {\n std::cerr << \"ggml_vulkan: Error: Vulkan 1.2 required.\" << std::endl;\n- GGML_ABORT(\"fatal error\");\n+ throw vk...
2025-09-27T16:26:46
ollama/ollama
22ab7f5f88c3a1f1704f8f5652417b91fd2a0245
2ecc3a33c380cd7915a8541d629d7f21c2b8472a
default host to `127.0.0.1`, fixes #424
[ { "path": "api/client.go", "patch": "@@ -16,7 +16,7 @@ import (\n \t\"github.com/jmorganca/ollama/version\"\n )\n \n-const DefaultHost = \"localhost:11434\"\n+const DefaultHost = \"127.0.0.1:11434\"\n \n var (\n \tenvHost = os.Getenv(\"OLLAMA_HOST\")", "additions": 1, "deletions": 1, "language":...
2023-08-26T18:59:28
denoland/deno
e5b1487f7f6dc18b8ce8618ba3f8455f8379ca8b
66c2ed6ed0c2c8a99388721dca8eca67b4ed5c7b
fix(webgpu): update wgpu to 28.0.0 (#31163)
[ { "path": "Cargo.lock", "patch": "@@ -96,7 +96,7 @@ dependencies = [\n \"cfg-if\",\n \"once_cell\",\n \"version_check\",\n- \"zerocopy\",\n+ \"zerocopy 0.7.32\",\n ]\n \n [[package]]\n@@ -717,18 +717,18 @@ dependencies = [\n \n [[package]]\n name = \"bytemuck\"\n-version = \"1.21.0\"\n+version = \"1.24.0...
2026-01-02T06:28:49
huggingface/transformers
927aa8bef2f29296a34840b3562f9c03cc45ef81
1951f3be8eb7f70afa408c61cb2c2f2c386d541a
[kernels] Cleanup deta kernel (#41470) * cleanup deta kernel * fix modeling
[ { "path": "src/transformers/kernels/deta/cpu/ms_deform_attn_cpu.cpp", "patch": "@@ -1,40 +0,0 @@\n-/*!\n-**************************************************************************************************\n-* Deformable DETR\n-* Copyright (c) 2020 SenseTime. All Rights Reserved.\n-* Licensed under the Apache...
2025-10-09T11:17:42
ggml-org/llama.cpp
72b24d96c6888c609d562779a23787304ae4609c
624207e676ab5eb3ce7af631902bb45fb73a8359
model : make minicpm embedding_scale, residual_scale and logit_scale optional with legacy defaults (#16273) * minicpm: make GGUF scaling keys optional with legacy defaults Older MiniCPM GGUFs do not include the scaling metadata keys (minicpm.embedding_scale, minicpm.residual_scale, minicpm.logit_scale). The loader cu...
[ { "path": "src/llama-model.cpp", "patch": "@@ -675,10 +675,17 @@ void llama_model::load_hparams(llama_model_loader & ml) {\n } break;\n case LLM_ARCH_MINICPM:\n {\n+ // Backward-compatible defaults for older MiniCPM GGUFs\n+ hparams.f_embedding_s...
2025-09-26T21:28:29
vuejs/vue
850fb9fbcb8990a66d15176085601f75cdb2e371
2d87b950c90adf9aec7564aeb3fe3f5e64fe1ef1
fix number update spec in ie9
[ { "path": "test/unit/features/directives/model-text.spec.js", "patch": "@@ -72,8 +72,8 @@ describe('Directive v-model text', () => {\n },\n template:\n '<div>' +\n- '<input ref=\"input\" v-model=\"test\" type=\"number\">{{ update }}' +\n- '<input ref=\"blur\"/>' +\n+ ...
2016-12-13T02:59:07
ollama/ollama
a9f6c56652dabbf77b64e695fb9bff6f0b6de797
0a892419ad377243f855053389fd26a42c431228
fix `FROM` instruction erroring when referring to a file
[ { "path": "cmd/cmd.go", "patch": "@@ -102,11 +102,15 @@ func RunHandler(cmd *cobra.Command, args []string) error {\n \t\treturn err\n \t}\n \n-\tmp, err := server.ParseModelPath(args[0], insecure)\n+\tmp := server.ParseModelPath(args[0])\n \tif err != nil {\n \t\treturn err\n \t}\n \n+\tif mp.ProtocolScheme...
2023-08-22T16:39:42
denoland/deno
66c2ed6ed0c2c8a99388721dca8eca67b4ed5c7b
d08d79083dd7b760aa556d057e4a937d4dd6c07e
fix: update urlpattern to 0.4.2 (#31193) Fixes #25953 Fixes #31072
[ { "path": "Cargo.lock", "patch": "@@ -4177,9 +4177,9 @@ checksum = \"aa9a19cbb55df58761df49b23516a86d432839add4af60fc256da840f66ed35b\"\n \n [[package]]\n name = \"form_urlencoded\"\n-version = \"1.2.1\"\n+version = \"1.2.2\"\n source = \"registry+https://github.com/rust-lang/crates.io-index\"\n-checksum = ...
2026-01-02T02:11:20
ggml-org/llama.cpp
624207e676ab5eb3ce7af631902bb45fb73a8359
807e8c6d310952f2f5656afc63dab9d7083dcb5c
devops: add s390x & ppc64le CI (#15925) * devops: move s390x and ppc64le ci build we have access to ubuntu-24.04-s390x and ppc64le images now Signed-off-by: Aaron Teo <aaron.teo1@ibm.com> * devops: disable ppc64le for now since they have compiler errors Signed-off-by: Aaron Teo <aaron.teo1@ibm.com> * devops: stop...
[ { "path": ".github/workflows/build-linux-cross.yml", "patch": "@@ -141,97 +141,6 @@ jobs:\n \n # cmake --build build --config Release -j $(nproc)\n \n- ubuntu-24-ppc64el-cpu-cross:\n- runs-on: ubuntu-24.04\n-\n- steps:\n- - uses: actions/checkout@v4\n- - name: Setup PowerPC64le\n-...
2025-09-26T18:03:33
vuejs/vue
2afa2601e05b51569d8d517624ffe0aa5e53a26a
6cfd6c7fe9c35528802d2e775833fb3eeebab0e8
ensure local assets is prioritized regardless of naming convention (fix #4434)
[ { "path": "src/core/util/options.js", "patch": "@@ -319,11 +319,14 @@ export function resolveAsset (\n return\n }\n const assets = options[type]\n- const res = assets[id] ||\n- // camelCase ID\n- assets[camelize(id)] ||\n- // Pascal Case ID\n- assets[capitalize(camelize(id))]\n+ // che...
2016-12-13T02:30:36
huggingface/transformers
be3fa93b298eec4182a0bb6c5e64270a5f1e8020
8137dbdbbd9c2f83b87aa9e6320b7a898f9d1073
Subconfig is a class attribute (#41308) * delete * fix this test * fix copies * oke, more tests to fix * fix last tests on DPT * deleted accidentally
[ { "path": "src/transformers/configuration_utils.py", "patch": "@@ -880,7 +880,6 @@ def to_diff_dict(self) -> dict[str, Any]:\n isinstance(getattr(self, key, None), PreTrainedConfig)\n and key in class_config_dict\n and isinstance(class_config_dict[key], dict)\...
2025-10-09T08:46:44
denoland/deno
d08d79083dd7b760aa556d057e4a937d4dd6c07e
0905fd9d36367fb46337c9be490a7337cffd758a
fix(ext/node): support numbered positional parameters in node:sqlite (#31722) SQLite numbered positional parameters (?1, ?2, ?NNN) were incorrectly treated as named parameters and skipped during binding, causing a "column index out of range" error. This fix updates the parameter binding logic to recognize that paramet...
[ { "path": "ext/node/ops/sqlite/statement.rs", "patch": "@@ -543,20 +543,33 @@ impl StatementSync {\n anon_start += 1;\n }\n \n- let mut anon_idx = 1;\n+ // SAFETY: `raw` is a valid pointer to a sqlite3_stmt.\n+ let sql_param_count = unsafe { ffi::sqlite3_bind_parameter_count(raw...
2026-01-01T16:04:36
ggml-org/llama.cpp
1a189278944d030211f336e103e96b65f976c361
e0539eb6aed346d4b25a6ea019044e88771e7690
Allow viewing conversations even when llama server is down (#16255) * webui: allow viewing conversations and sending messages even if llama-server is down - Cached llama.cpp server properties in browser localStorage on startup, persisting successful fetches and reloading them when refresh attempts fail so the chat UI...
[ { "path": "tools/server/webui/src/lib/components/app/chat/ChatScreen/ChatScreen.svelte", "patch": "@@ -3,9 +3,11 @@\n \timport {\n \t\tChatForm,\n \t\tChatScreenHeader,\n+\t\tChatScreenWarning,\n \t\tChatMessages,\n \t\tChatProcessingInfo,\n \t\tEmptyFileAlertDialog,\n+\t\tServerErrorSplash,\n \t\tServerInf...
2025-09-26T16:35:42
vuejs/vue
92473853dd70b6d38d88686c15c338f262da40e3
183bd43c3ee276aac9052a5eaae8b782cd99a095
avoid error when asserting mismatched nodes during hydration
[ { "path": "src/core/vdom/patch.js", "patch": "@@ -534,7 +534,7 @@ export function createPatchFunction (backend) {\n if (vnode.tag) {\n return (\n vnode.tag.indexOf('vue-component') === 0 ||\n- vnode.tag.toLowerCase() === node.tagName.toLowerCase()\n+ vnode.tag.toLowerCase() =...
2016-12-12T19:22:12
ollama/ollama
5ca05c2e88c8d605b33bafc241fc84b77d322fc4
7eda70f23b08854a341542839c43e831acc976d1
fix ModelType()
[ { "path": "llm/llama.go", "patch": "@@ -117,7 +117,21 @@ func (llm *llamaModel) ModelFamily() ModelFamily {\n }\n \n func (llm *llamaModel) ModelType() ModelType {\n-\treturn ModelType30B\n+\tswitch llm.hyperparameters.NumLayer {\n+\tcase 26:\n+\t\treturn ModelType3B\n+\tcase 32:\n+\t\treturn ModelType7B\n+...
2023-08-18T18:23:31
huggingface/transformers
7aa888b7fa477d13153ffbfe107dfbd6c696014a
bfe2b623ef9bd301e57463258ffcc5e6fb19e2ac
Fix doc (#41457) * dummy * remove
[ { "path": "docs/source/ko/internal/generation_utils.md", "patch": "@@ -252,8 +252,6 @@ generation_output[:2]\n [[autodoc]] SlidingWindowCache\n \n [[autodoc]] EncoderDecoderCache\n- - to_legacy_cache\n- - from_legacy_cache\n \n ## 워터마크 유틸리티 (Watermark Utils) [[transformers.WatermarkDetector]]\n ", ...
2025-10-08T18:13:21
ggml-org/llama.cpp
e0539eb6aed346d4b25a6ea019044e88771e7690
5d0a40f390732cbf85d8a3b7b0fc3cbebffe780a
webui: switch to hash-based routing (alternative of #16079) (#16157) * Switched web UI to hash-based routing * Added hash to missed goto function call * Removed outdated SPA handling code * Fixed broken sidebar home link
[ { "path": "tools/server/server.cpp", "patch": "@@ -5262,42 +5262,6 @@ int main(int argc, char ** argv) {\n svr->Get (params.api_prefix + \"/slots\", handle_slots);\n svr->Post(params.api_prefix + \"/slots/:id_slot\", handle_slots_action);\n \n- // SPA fallback route - serve ind...
2025-09-26T15:36:48
denoland/deno
0905fd9d36367fb46337c9be490a7337cffd758a
0ffed3e8fbd1ff0c26f37539e20401ddb53d3119
feat(node/http): enable keepAlive connection reuse for HTTP Agent (#31709) Closes https://github.com/denoland/deno/issues/29676 Enables HTTP Agent keepAlive connection pooling for improved performance with Node.js HTTP clients. Changes: - Use `conn.without_shutdown()` to allow socket reuse after HTTP response - Add ...
[ { "path": "ext/node/lib.rs", "patch": "@@ -323,6 +323,7 @@ deno_core::extension!(deno_node,\n ops::handle_wrap::op_node_new_async_id,\n ops::http::op_node_http_fetch_response_upgrade,\n ops::http::op_node_http_request_with_conn,\n+ ops::http::op_node_http_response_reclaim_conn,\n ops::htt...
2025-12-30T13:22:08
vuejs/vue
183bd43c3ee276aac9052a5eaae8b782cd99a095
962b778e109807c2da34025c9ca8f6ba08c48bbe
fix v-bind.prop parse (fix #4432) (#4435) * reset isProp value * add test case * update test case * fix typo
[ { "path": "src/compiler/parser/index.js", "patch": "@@ -414,6 +414,7 @@ function processAttrs (el) {\n if (bindRE.test(name)) { // v-bind\n name = name.replace(bindRE, '')\n value = parseFilters(value)\n+ isProp = false\n if (modifiers) {\n if (modifiers.prop) ...
2016-12-11T15:16:23
ggml-org/llama.cpp
cc1cfa277b3aae1f6cc9180472072336597a78d4
54dbc37053f7d75ea5b0631d5ee88f19391e4314
mtmd : fix uninitialized variable in bicubic_resize (#16275) Signed-off-by: Aaron Teo <aaron.teo1@ibm.com> Co-authored-by: Aaron Teo <aaron.teo1@ibm.com>
[ { "path": "tools/mtmd/clip.cpp", "patch": "@@ -3067,7 +3067,7 @@ struct image_manipulation {\n dst.buf.resize(3 * target_width * target_height);\n \n float Cc;\n- float C[5];\n+ float C[5] = {};\n float d0, d2, d3, a0, a1, a2, a3;\n int i, j, k, jj;\n in...
2025-09-26T13:00:44
huggingface/transformers
bfe2b623ef9bd301e57463258ffcc5e6fb19e2ac
b9be8a8775ff0899912f6db126fa8c770b6698cd
Fix generate outputs and simplify cache tests (#41440) * start refactoring * simplify * tests * tests * fix * zamba * final fix * fix
[ { "path": "src/transformers/generation/utils.py", "patch": "@@ -2812,6 +2812,10 @@ def _sample(\n streamer.end()\n \n if return_dict_in_generate:\n+ cache = None\n+ if any(cache_key in model_kwargs for cache_key in ALL_CACHE_NAMES):\n+ cache_key = nex...
2025-10-08T17:04:18
ollama/ollama
3cbc6a5c01120b59cf38378d5f18836bcd90f6c9
54bb49a50264270e97ade94082691859668f99ee
fix push manifest
[ { "path": "server/images.go", "patch": "@@ -963,18 +963,12 @@ func PushModel(ctx context.Context, name string, regOpts *RegistryOptions, fn fu\n \t\treturn err\n \t}\n \n-\tresp, err := makeRequest(ctx, \"PUT\", url, headers, bytes.NewReader(manifestJSON), regOpts)\n+\tresp, err := makeRequestWithRetry(ctx,...
2023-08-17T19:35:29
denoland/deno
0ffed3e8fbd1ff0c26f37539e20401ddb53d3119
0ae83e43ca96a1b40e7c8bcb80590f6a1fe757c2
fix(ext/node): support TypedArray and DataView targets in Buffer.copy() (#31734) Closes https://github.com/denoland/deno/issues/31637 - `parallel/test-buffer-copy.js` now passes - Allow `Buffer.copy()` to accept any `ArrayBufferView` (TypedArray or DataView) as the target, matching Node.js behavior - Previously only ...
[ { "path": "ext/node/polyfills/internal/buffer.mjs", "patch": "@@ -17,7 +17,9 @@ const {\n ArrayPrototypeSlice,\n ArrayPrototypeForEach,\n BigInt,\n+ DataViewPrototypeGetBuffer,\n DataViewPrototypeGetByteLength,\n+ DataViewPrototypeGetByteOffset,\n Float32Array,\n Float64Array,\n MathFloor,\n...
2025-12-30T13:19:49
vuejs/vue
962b778e109807c2da34025c9ca8f6ba08c48bbe
29d6e33b33a6199f15f574d6d9192176713718b6
More informative warning message for comp name (#4429) This commit adds a more informative warning message for invalid component names. Also fixes a typo. Closes #4428.
[ { "path": "src/core/global-api/extend.js", "patch": "@@ -28,7 +28,8 @@ export function initExtend (Vue: GlobalAPI) {\n if (!/^[a-zA-Z][\\w-]*$/.test(name)) {\n warn(\n 'Invalid component name: \"' + name + '\". Component names ' +\n- 'can only contain alphanumeric characater...
2016-12-09T17:51:13
huggingface/transformers
b9be8a8775ff0899912f6db126fa8c770b6698cd
bef73bf8d7421e2158a83ebcc24a9fc78e885dfe
enable some falcon-mamba uts on xpu (#41428) * enable some falcon-mamba uts on xpu Signed-off-by: Yao, Matrix <matrix.yao@intel.com> * fix style Signed-off-by: Yao, Matrix <matrix.yao@intel.com> --------- Signed-off-by: Yao, Matrix <matrix.yao@intel.com>
[ { "path": "src/transformers/models/falcon_mamba/modeling_falcon_mamba.py", "patch": "@@ -170,8 +170,13 @@ def _lazy_load_causal_conv1d():\n if is_kernels_available():\n from kernels import get_kernel\n \n- _causal_conv1d_kernel = get_kernel(\"kernels-community/causal-conv1d\")\n- _...
2025-10-08T16:48:04
ggml-org/llama.cpp
54dbc37053f7d75ea5b0631d5ee88f19391e4314
b995a10760cb93d23d617d76ecb82a5f95b5e0d3
metal : report OOM errors (#16274)
[ { "path": "ggml/src/ggml-metal/ggml-metal-context.m", "patch": "@@ -222,7 +222,28 @@ void ggml_metal_synchronize(ggml_metal_t ctx) {\n ctx->cmd_buf_last = nil;\n }\n \n- // release any completed command buffers\n+ // check status of all command buffers\n+ {\n+ const int n_cb = ct...
2025-09-26T11:14:28
denoland/deno
0ae83e43ca96a1b40e7c8bcb80590f6a1fe757c2
223d7bb683a425ecbf5a6380a9a0a5fd46581a20
fix(task): prioritize member `package.json` when `deno.json` only in root (#31736) It was using the deno.json from the root even if the member package.json had a script with the same name.
[ { "path": "cli/args/mod.rs", "patch": "@@ -799,7 +799,7 @@ impl CliOptions {\n pub fn resolve_storage_key_resolver(&self) -> StorageKeyResolver {\n if let Some(location) = &self.flags.location {\n StorageKeyResolver::from_flag(location)\n- } else if let Some(deno_json) = self.start_dir.maybe_...
2025-12-29T22:49:03
vuejs/vue
612d32105a4ba1d636e8b617cbcb0a3e59919a9e
0163a6fe53cc17a69450f5c0ef0ac16308c1f5b6
fix transition v-show display toggle timing for enter hooks (fix #4418)
[ { "path": "src/platforms/web/runtime/directives/show.js", "patch": "@@ -14,22 +14,28 @@ export default {\n bind (el: any, { value }: VNodeDirective, vnode: VNodeWithData) {\n vnode = locateNode(vnode)\n const transition = vnode.data && vnode.data.transition\n+ const originalDisplay = el.__vOrig...
2016-12-08T21:54:09
ggml-org/llama.cpp
4710dd31bbcef79d04f85a3a6a8c7d9439c5c79a
9b26511857ac09ae69ab485168fe2d3ee5fb1d6e
build : fix build-ios-device (#16257) Signed-off-by: Adrien Gallouët <angt@huggingface.co>
[ { "path": "build-xcframework.sh", "patch": "@@ -422,6 +422,7 @@ echo \"Building for iOS devices...\"\n cmake -B build-ios-device -G Xcode \\\n \"${COMMON_CMAKE_ARGS[@]}\" \\\n -DCMAKE_OSX_DEPLOYMENT_TARGET=${IOS_MIN_OS_VERSION} \\\n+ -DCMAKE_SYSTEM_NAME=iOS \\\n -DCMAKE_OSX_SYSROOT=iphoneos \...
2025-09-26T10:39:35
denoland/deno
223d7bb683a425ecbf5a6380a9a0a5fd46581a20
ce1d4514ac401d8550df16b59b37c20bb5958b3b
fix(serve): net permissions in config file (#31718) It wasn't being resolved properly because we were injecting allow-net options when parsing flags instead of when resolving wholly resolving permissions. Closes #31728
[ { "path": "cli/args/flags.rs", "patch": "@@ -6509,25 +6509,6 @@ fn serve_parse(\n let open_site = matches.remove_one::<bool>(\"open\").unwrap_or(false);\n \n runtime_args_parse(flags, matches, true, true, true)?;\n- // If the user didn't pass --allow-net, add this port to the network\n- // allowlist. ...
2025-12-29T14:11:22
huggingface/transformers
89a4115a6b3697a3a2018d8d4f8c439e4196691e
82ffeb28ad926938db1f81e2423b6ba4ffbed579
Validate processing kwargs with @strict from huggingface_hub (#40793) * initial design draft * delete * fix a few tests * fix * fix the rest of tests * common-kwargs * why the runner complains about typing with "|"? * revert * forgot to delete * update * fix last issues * add more detalis in docs * pin th...
[ { "path": "setup.py", "patch": "@@ -114,7 +114,7 @@\n \"GitPython<3.1.19\",\n \"hf-doc-builder>=0.3.0\",\n \"hf_xet\",\n- \"huggingface-hub==1.0.0.rc2\",\n+ \"huggingface-hub==1.0.0.rc4\",\n \"importlib_metadata\",\n \"ipadic>=1.0.0,<2.0\",\n \"jinja2>=3.1.0\",", "additions...
2025-10-08T14:14:09