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
97508acb17ff933c67edb3a2f0997bc19a6abb98
5c4aae66e15990f87815c9eba1663f728067512b
webui: fix syntax highlighting lost after streaming for non-common languages (#21206) * webui: fix syntax highlighting lost for non-common languages after streaming rehype-highlight uses lowlight internally, which only bundles 37 "common" languages. The streaming code path uses highlight.js directly (192 languages), ...
[ { "path": "tools/server/public/index.html", "patch": "@@ -18,7 +18,7 @@\n \t\t<div style=\"display: contents\">\n \t\t\t<script>\n \t\t\t\t{\n-\t\t\t\t\t__sveltekit_10avopp = {\n+\t\t\t\t\t__sveltekit_1ppa22i = {\n \t\t\t\t\t\tbase: new URL('.', location).pathname.slice(0, -1)\n \t\t\t\t\t};\n ", "addit...
2026-04-08T06:58:08
golang/go
38367d098ed4d97539de5e43e03bce985fc56d8e
d834ecec8637e3d54b67debf95ceb649cc0b4e1d
cmd/link/internal/ld: dedup shared libraries on openbsd When linking internally on OpenBSD, dedup libraries treating versioned and unversioned libraries as equivalents. Versioned libraries are preferred and are retained over unversioned libraries. This avoids the situation where the use of cgo results in a DT_NEEDED ...
[ { "path": "src/cmd/link/internal/ld/elf_test.go", "patch": "@@ -14,6 +14,7 @@ import (\n \t\"os/exec\"\n \t\"path/filepath\"\n \t\"runtime\"\n+\t\"strings\"\n \t\"testing\"\n )\n \n@@ -123,7 +124,7 @@ func TestNoDuplicateNeededEntries(t *testing.T) {\n \n \tvar count int\n \tfor _, lib := range libs {\n-\t\...
2020-11-11T17:30:15
nodejs/node
7db53706c9aa44aa645a410cd48cf23cf1f300fd
bbcbcb810b08b1b8936c8f64648edfae35beaa62
crypto: do not reach into OpenSSL internals for ThrowCryptoError There is a perfectly serviceable ERR_get_error function which avoids having to sniff through the OpenSSL ring buffer like that. It does return the errors in the opposite order, but that's easily fixed with std::reverse. Note this behavior is slightly di...
[ { "path": "src/node_crypto.cc", "patch": "@@ -42,11 +42,13 @@\n // StartComAndWoSignData.inc\n #include \"StartComAndWoSignData.inc\"\n \n+#include <algorithm>\n #include <errno.h>\n #include <limits.h> // INT_MAX\n #include <math.h>\n #include <stdlib.h>\n #include <string.h>\n+#include <vector>\n \n #def...
2017-11-03T00:09:31
rust-lang/rust
c62aa0baa1a8228d5bfbb3e810db4c7ee77eb3a1
8c7a94e4cdd9ad70550b19ca2bf6f16046d59506
Fix `UserRef<[T]>::copy_to_enclave_vec` It reinterprets uninitialized memory as initialized and does not drop existing elements of the Vec. Fix that. Additionally, make it more general by appending, instead of overwriting existing elements, and rename it to `append_to_enclave_vec`. A caller can simply call `.clear()`...
[ { "path": "library/std/src/sys/pal/sgx/abi/usercalls/alloc.rs", "patch": "@@ -678,25 +678,18 @@ where\n unsafe { (*self.0.get()).len() }\n }\n \n- /// Copies the value from user memory and place it into `dest`. Afterwards,\n- /// `dest` will contain exactly `self.len()` elements.\n- ///...
2025-02-26T19:51:28
huggingface/transformers
5019aabfacf7599b9a6b4e7a1adc1fb5c9017727
f2122cc6eb8e50e4d1b45da54b43bba59a458b30
Optimize t5 tokenize logic to avoid redundant calls (#32270) * Optimize t5 tokenize logic to avoid redundant calls * fix and overwrite copies
[ { "path": "src/transformers/models/llama/tokenization_llama.py", "patch": "@@ -261,9 +261,8 @@ def _tokenize(self, text, **kwargs):\n `unk_token`. Here is an example with `unk_token = \"<unk>\"` and `unk_token_length = 4`.\n `self.tokenizer.sp_model.encode(\"<unk> Hey\", out_type = str)[4:]`...
2024-07-29T07:51:43
ggml-org/llama.cpp
66c4f9ded01b29d9120255be1ed8d5835bcbb51d
93bdc6156333082421d82b52d36f7dacb0542495
ggml-cuda: ds_read_b128 for q4_0 and q4_1 mmq kernels (#21168) * ds_read_b128 for q4_0 and q4_1 mmq kernels Current for loop generates ds_read_b32 instructions with hip compiler, the new solution generates ds_read_b128 instructions for the same operation, saving some LDS bandwidth. Tested on MI50 and RX6800XT, i...
[ { "path": "ggml/src/ggml-cuda/mmq.cuh", "patch": "@@ -386,17 +386,25 @@ static __device__ __forceinline__ void vec_dot_q4_0_q8_1_dp4a(\n #pragma unroll\n for (int i0 = 0; i0 < mmq_y; i0 += warp_size) {\n const int i = i0 + threadIdx.x;\n-\n const int kyqs = QI8_1 ...
2026-04-07T19:47:42
golang/go
d834ecec8637e3d54b67debf95ceb649cc0b4e1d
0932dc21180642ce1ff095b9b3e68b06c6f440b3
runtime/race: reject runtime fatal error in tests We expect those tests to fail with non-zero exit code, due to intentional races, but we don't expect the runtime to crash. Reject that. Change-Id: Ic37987dabecde5f0703c031c49ce7f884a7b06a5 Reviewed-on: https://go-review.googlesource.com/c/go/+/270398 Trust: Cherry Zha...
[ { "path": "src/runtime/race/race_test.go", "patch": "@@ -177,6 +177,10 @@ func runTests(t *testing.T) ([]byte, error) {\n \t)\n \t// There are races: we expect tests to fail and the exit code to be non-zero.\n \tout, _ := cmd.CombinedOutput()\n+\tif bytes.Contains(out, []byte(\"fatal error:\")) {\n+\t\t// B...
2020-11-14T19:29:31
nodejs/node
86527bd762576149bd954fff055954d2e43a05ce
e9d1e1265a8d8aa708ae2a57088e43d08c81415e
util: fix negative 0 check in inspect PR-URL: https://github.com/nodejs/node/pull/17507 Reviewed-By: Refael Ackermann <refack@gmail.com> Reviewed-By: Anatoli Papirovski <apapirovski@mac.com> Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl> Reviewed-By: Colin Ihrig <cjihrig@gmail.com> Reviewed-By: Khaidi Chu <i@2333.moe...
[ { "path": "lib/util.js", "patch": "@@ -597,10 +597,8 @@ function formatValue(ctx, value, recurseTimes, ln) {\n }\n \n function formatNumber(fn, value) {\n- // Format -0 as '-0'. A `value === -0` check won't distinguish 0 from -0.\n- // Using a division check is currently faster than `Object.is(value, -0)`...
2017-12-06T21:09:07
ggml-org/llama.cpp
93bdc6156333082421d82b52d36f7dacb0542495
4eb19514dd2984662f13aacbb052c559c8fde3b1
gguf-py : fix missing comma after bad merge in tensor-mapping (#21558) This commit adds a missing comma in the vision encoder attention qkv block. The motivation for this change is that without the comma there will be a string concatenation of the Kimi-K2.5 and the Nemotron Nano v2 VL tensor mappings which will be br...
[ { "path": "gguf-py/gguf/tensor_mapping.py", "patch": "@@ -1441,7 +1441,7 @@ class TensorNameMap:\n \"visual.blocks.{bid}.attn.qkv\", # qwen3vl\n \"model.vision.transformer.layers.{bid}.attention.query_key_value\", # cogvlm\n \"model.vision_model.transformer.layers.{bid}.s...
2026-04-07T19:24:25
huggingface/transformers
f7396876849926afa87c9412d67c43618dad403d
44f6fdd74f84744b159fa919474fd3108311a906
🚨 Bloom support for cache class (#31445) * bloom dynamic cache * bloom follows standard cache format * no skips for bloom anymore * use cache position when possible * clean up * codestyle * Update src/transformers/models/bloom/modeling_bloom.py Co-authored-by: amyeroberts <22614925+amyeroberts@us...
[ { "path": "src/transformers/generation/candidate_generator.py", "patch": "@@ -378,19 +378,7 @@ def _crop_past_key_values(model, past_key_values, max_length):\n )\n )\n past_key_values = tuple(new_past)\n- # bloom is special\n- elif \"bloom\" in model.__class__.__nam...
2024-07-29T05:58:59
golang/go
0932dc21180642ce1ff095b9b3e68b06c6f440b3
d70a33a40bd2bab2f8cd6ab714c4664ce55dc499
runtime: declare arg size/map for race version of sync/atomic functions The argument size and map are used in stack scanning if those functions are deferred. Declare the right argument size and map so they can be scanned correctly. Fixes #42599. Change-Id: I74f9409d574cf7c383f4d8f83e38521026b48861 Reviewed-on: https...
[ { "path": "src/runtime/race/testdata/atomic_test.go", "patch": "@@ -299,3 +299,27 @@ func TestNoRaceAtomicCrash(t *testing.T) {\n \t}()\n \tatomic.AddInt32(nilptr, 1)\n }\n+\n+func TestNoRaceDeferAtomicStore(t *testing.T) {\n+\t// Test that when an atomic function is deferred directly, the\n+\t// GC scans i...
2020-11-14T02:08:26
ollama/ollama
4ad0d4d6d3bd9b3507644d872fb27113fd3e487b
163cd3e77c42aafd003b9cb884b3a51cdbaea106
Fix a build warning (#5096) Signed-off-by: Lei Jitang <leijitang@outlook.com>
[ { "path": "gpu/gpu_info_oneapi.c", "patch": "@@ -98,7 +98,7 @@ void oneapi_init(char *oneapi_lib_path, oneapi_init_resp_t *resp) {\n }\n \n for (d = 0; d < resp->oh.num_drivers; d++) {\n- LOG(resp->oh.verbose, \"calling zesDeviceGet %d\\n\", resp->oh.drivers[d]);\n+ LOG(resp->oh.verbose, \"calling...
2024-06-17T18:47:48
ggml-org/llama.cpp
957d717ce54401a28591c265d19c53ba620f4c46
de1aa6fa73e135839109e09fec1a0997f4207b2a
ggml-webgpu: parameterize submission size and add iOS specific limits (#21533) * Work towards removing bitcast * Move rest of existing types over * Add timeout back to wait and remove synchronous set_tensor/memset_tensor * move to unpackf16 for wider compatibility * cleanup * Remove deadlock condition in free_buf...
[ { "path": "ggml/src/ggml-webgpu/ggml-webgpu.cpp", "patch": "@@ -16,7 +16,6 @@\n #include <webgpu/webgpu_cpp.h>\n \n #include <atomic>\n-#include <condition_variable>\n #include <cstdint>\n #include <cstring>\n #ifdef GGML_WEBGPU_GPU_PROFILE\n@@ -25,7 +24,6 @@\n #if defined(GGML_WEBGPU_DEBUG) || defined(GGML...
2026-04-07T17:30:01
nodejs/node
df79b7d8214f61df1520c15f0b44d7ef7808d345
d865395b7d81822acbeae5e7f3b95eb3a357a37f
src: fix missing handlescope bug in inspector Fix a regression that was introduced in commit 5886e204f0 ("inspector: track async stacks when necessary") and that I overlooked during review: the persistent handle with the callback must be rematerialized *after* the `v8::HandleScope` is created, not before. Apparently ...
[ { "path": "src/inspector_agent.cc", "patch": "@@ -30,6 +30,7 @@ using v8::HandleScope;\n using v8::Isolate;\n using v8::Local;\n using v8::Object;\n+using v8::Persistent;\n using v8::Value;\n \n using v8_inspector::StringBuffer;\n@@ -613,8 +614,7 @@ void Agent::RegisterAsyncHook(Isolate* isolate,\n \n void ...
2017-12-07T22:55:23
huggingface/transformers
81233c069c166af033794134bd8888783ac49ebe
27c7f971c0dcd3bb423ea221fe2bce751d313119
Flash-Attn: fix generation when no attention mask or no pading (#32241) * fix * fix prev test (half of failures) * [run-slow] llama, gemma2 * [run-slow] llama, gemma2
[ { "path": "src/transformers/modeling_flash_attention_utils.py", "patch": "@@ -264,9 +264,11 @@ def _flash_attention_forward(\n )\n attn_output = pad_input(attn_output_unpad, indices_q, batch_size, query_length)\n \n- # if position_ids is provided and check not all examples (row) contain o...
2024-07-26T09:45:55
rust-lang/rust
7398b39a0bfae6c778379b46f2d9e957c94f4f0d
64d143326f222ef477044d846e7e8f09ac6d6a3a
Make panic's more specific
[ { "path": "compiler/rustc_index/src/slice.rs", "patch": "@@ -1,6 +1,7 @@\n use std::fmt;\n use std::marker::PhantomData;\n use std::ops::{Index, IndexMut};\n+use std::slice::GetDisjointMutError::*;\n use std::slice::{self, SliceIndex};\n \n use crate::{Idx, IndexVec, IntoSliceIdx};\n@@ -115,33 +116,35 @@ im...
2025-03-11T22:35:16
golang/go
af814af6e7b48e0aa2784203297c0ba69fe9de5c
f42bd50779dea8d8e46de14e2f00cfe716f52d6d
[dev.boringcrypto] misc/boring: add new releases to RELEASES file Fixes #42589 Change-Id: Ieba62381d6561d4803448c123ce1f1d5980b8a76 Reviewed-on: https://go-review.googlesource.com/c/go/+/269762 Trust: Filippo Valsorda <filippo@golang.org> Reviewed-by: Katie Hockman <katie@golang.org>
[ { "path": "misc/boring/RELEASES", "patch": "@@ -105,3 +105,7 @@ go1.14.10b4 b5fc12785be4 linux-amd64 https://go-boringcrypto.storage.googleapis.\n go1.14.10b4 b5fc12785be4 src https://go-boringcrypto.storage.googleapis.com/go1.14.10b4.src.tar.gz daf7603babc49935efdea5befb2ecad823771523a84d1ba6c0e8c10fac982d...
2020-11-13T17:04:13
ollama/ollama
163cd3e77c42aafd003b9cb884b3a51cdbaea106
4c2c8f93ddcb41e362ad97306cd83b38b5449e56
gpu: add env var for detecting Intel oneapi gpus (#5076) * gpu: add env var for detecting intel oneapi gpus * fix build error
[ { "path": "envconfig/config.go", "patch": "@@ -57,6 +57,8 @@ var (\n \tSchedSpread bool\n \t// Set via OLLAMA_TMPDIR in the environment\n \tTmpDir string\n+\t// Set via OLLAMA_INTEL_GPU in the environment\n+\tIntelGpu bool\n \n \t// Set via CUDA_VISIBLE_DEVICES in the environment\n \tCudaVisibleDevices stri...
2024-06-17T00:09:05
ggml-org/llama.cpp
69c28f1547c169902f62ca48bee75fb876c4d8e6
0d049d6a9245ccffc6073743ff8b6bb24ac6a47b
llama-server: fix model params not propagated (#21509) Signed-off-by: Aaron Teo <aaron.teo1@ibm.com>
[ { "path": "tools/server/server-context.cpp", "patch": "@@ -632,7 +632,7 @@ struct server_context_impl {\n \n // load the model and initialize llama_context\n // this may also be called to resume from sleeping state\n- bool load_model(const common_params & params) {\n+ bool load_model(common_pa...
2026-04-07T13:39:41
nodejs/node
f01e9e7f6961ad1bde488d2780e96f4bea32f9dd
762ed339319adffd62cdd9d0eda8bb3fb6c73f86
test: fix flaky test-benchmark-es Allow zero iterations for short benchmark in test. PR-URL: https://github.com/nodejs/node/pull/17516 Reviewed-By: Anatoli Papirovski <apapirovski@mac.com> Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
[ { "path": "test/parallel/test-benchmark-es.js", "patch": "@@ -15,4 +15,5 @@ runBenchmark('es',\n 'n=1',\n 'encoding=ascii',\n 'size=1e1'\n- ]);\n+ ],\n+ { NODEJS_BENCHMARK_ZERO_ALLOWED: 1 });", "additions": 2, "deletion...
2017-12-07T03:26:36
huggingface/transformers
27c7f971c0dcd3bb423ea221fe2bce751d313119
5f841c74b62754f186a8c06a684d491524b7bc03
[tests] fix `static` cache implementation is not compatible with `attn_implementation==flash_attention_2` (#32039) * add flash attention check * fix * fix
[ { "path": "tests/utils/test_cache_utils.py", "patch": "@@ -290,7 +290,7 @@ def test_sink_cache_iterative_prompts(self):\n self.assertTrue(decoded[0].endswith(last_output))\n \n @require_torch_gpu\n- @parameterized.expand([\"eager\", \"sdpa\", \"flash_attention_2\"])\n+ @parameterized.expan...
2024-07-26T09:41:27
rust-lang/rust
763db5dcd9b95179915e13533af280ad5f6db403
6650252439d4e03368b305c42a10006e36f1545e
Convert a delayed bug to a bug. This is never hit in the test suite. At some point the check should be removed entirely. There are a million places in the compiler where an empty symbol doesn't make sense, so a check of this nature has almost zero value. But I'll leave it in place for now just in case it gets hit by ...
[ { "path": "compiler/rustc_hir_typeck/src/expr.rs", "patch": "@@ -2933,7 +2933,7 @@ impl<'a, 'tcx> FnCtxt<'a, 'tcx> {\n }\n \n let guar = if field.name == kw::Empty {\n- self.dcx().span_delayed_bug(field.span, \"field name with no name\")\n+ self.dcx().span_bug(field.spa...
2025-03-05T06:11:46
ggml-org/llama.cpp
0d049d6a9245ccffc6073743ff8b6bb24ac6a47b
a8ec0df4617692ef0f18b212f2e16bba540e053a
unicode : add custom Qwen2 regex handler to fix segfault on long input (#21257) * unicode : add custom Qwen2 regex handler to fix segfault on long input std::regex uses recursive backtracking internally, which causes a stack overflow (segfault) when tokenizing long sequences of repeated characters (e.g. 43K 'A's). Th...
[ { "path": "src/unicode.cpp", "patch": "@@ -470,6 +470,141 @@ static std::vector<size_t> unicode_regex_split_custom_llama3(const std::string &\n return bpe_offsets;\n }\n \n+// Qwen2 system regex: \"(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\\\r\\\\n\\\\p{L}\\\\p{N}]?\\\\p{L}+|\\\\p{N}| ?[^\\\\s\\\\p{L}\\\\p{N}]+[...
2026-04-07T13:13:38
golang/go
c7233dd063cd8bf24460b280e3929458e64f4315
f2eea4c1dc37886939c010daff89c03d5a3825be
cmd/go: permit wrongly rejected -Wl,-O... linker flags A typo caused the validation rule to check against -WL,-O... which is not a regular flag because the L should be lowercase as in the other rules. This caused valid linker flags to be rejected and people had to work around this by filtering their default flags that...
[ { "path": "src/cmd/go/internal/work/security.go", "patch": "@@ -179,7 +179,7 @@ var validLinkerFlags = []*lazyregexp.Regexp{\n \tre(`-Wl,-berok`),\n \tre(`-Wl,-Bstatic`),\n \tre(`-Wl,-Bsymbolic-functions`),\n-\tre(`-WL,-O([^@,\\-][^,]*)?`),\n+\tre(`-Wl,-O([^@,\\-][^,]*)?`),\n \tre(`-Wl,-d[ny]`),\n \tre(`-Wl...
2020-11-16T12:05:13
huggingface/transformers
5f841c74b62754f186a8c06a684d491524b7bc03
f9756d9edb23354e3df50f7eb3f6b3129a25e453
Add check for `target_sizes is None` in `post_process_image_guided_detection` for owlv2 (#31934) * Add check for target_sizes is None in post_process_image_guided_detection * Make sure Owlvit and Owlv2 in sync * Fix incorrect indentation; add check for correct size of target_sizes
[ { "path": "src/transformers/models/owlv2/image_processing_owlv2.py", "patch": "@@ -565,9 +565,9 @@ def post_process_image_guided_detection(self, outputs, threshold=0.0, nms_thresh\n \"\"\"\n logits, target_boxes = outputs.logits, outputs.target_pred_boxes\n \n- if len(logits) != len(t...
2024-07-26T09:05:46
nodejs/node
146a9b1a8a38285b0613559aa96498bc7859fc75
d8c896cac5b3e31cbffb3da2c657dd59c83fbfeb
test: replace assert.throws w/ common.expectsError PR-URL: https://github.com/nodejs/node/pull/17483 Reviewed-By: Ruben Bridgewater <ruben@bridgewater.de> Reviewed-By: Colin Ihrig <cjihrig@gmail.com> Reviewed-By: Evan Lucas <evanlucas@me.com> Reviewed-By: James M Snell <jasnell@gmail.com>
[ { "path": "test/parallel/test-async-hooks-asyncresource-constructor.js", "patch": "@@ -3,7 +3,6 @@\n // This tests that AsyncResource throws an error if bad parameters are passed\n \n const common = require('../common');\n-const assert = require('assert');\n const async_hooks = require('async_hooks');\n con...
2017-12-06T08:10:46
rust-lang/rust
650b7d957b463bd760cfa7fdbc42619c172a7626
cb044d4d7b6d4b9f944a648b38c790e6fadae9b8
Move methods from `Map` to `TyCtxt`, part 4. Continuing the work from #137350. Removes the unused methods: `expect_variant`, `expect_field`, `expect_foreign_item`. Every method gains a `hir_` prefix.
[ { "path": "clippy_lints/src/attrs/mod.rs", "patch": "@@ -465,7 +465,7 @@ impl Attributes {\n \n impl<'tcx> LateLintPass<'tcx> for Attributes {\n fn check_item(&mut self, cx: &LateContext<'tcx>, item: &'tcx Item<'_>) {\n- let attrs = cx.tcx.hir().attrs(item.hir_id());\n+ let attrs = cx.tcx....
2025-02-21T07:33:05
ggml-org/llama.cpp
e8f50826979b5d3d4f54e522e552caf430edce0d
22fc79134e23feea5cbc1074eaae77143932ccbc
server : fix restore for checkpoints with pos_min == 0 (#21510)
[ { "path": "tools/server/server-context.cpp", "patch": "@@ -2404,7 +2404,7 @@ struct server_context_impl {\n // guarantee that a checkpoint will result in at least one token being processed [TAG_PROMPT_LOGITS]\n LOG_INF(\...
2026-04-07T12:29:17
ollama/ollama
fd1e6e0590894b795a5b041736a43bc1498adc22
89c79bec8cf7a7b04f761fcc5306d2edf47a4164
Add some more debugging logs for intel discovery Also removes an unused overall count variable
[ { "path": "gpu/gpu_info_oneapi.c", "patch": "@@ -13,7 +13,7 @@ void oneapi_init(char *oneapi_lib_path, oneapi_init_resp_t *resp) {\n resp->oh.num_drivers = 0;\n const int buflen = 256;\n char buf[buflen + 1];\n- int i, d, count;\n+ int i, d;\n struct lookup {\n char *s;\n void **p;\n@@ -62...
2024-06-16T14:42:52
golang/go
f2eea4c1dc37886939c010daff89c03d5a3825be
92c732e901a732855f4b813e6676264421eceae9
cmd/compile: mask SLL,SRL,SRAconst shift amount mips SRA/SLL/SRL shift amounts are used mod 32; this change aligns the XXXconst rules to mask the shift amount by &31. Passes $ GOARCH=mips go build -toolexec 'toolstash -cmp' -a std $ GOARCH=mipsle go build -toolexec 'toolstash -cmp' -a std Fixes #42587 Change-I...
[ { "path": "src/cmd/compile/internal/ssa/gen/MIPS.rules", "patch": "@@ -567,10 +567,9 @@\n (XOR x (MOVWconst [c])) => (XORconst [c] x)\n (NOR x (MOVWconst [c])) => (NORconst [c] x)\n \n-(SRA x (MOVWconst [c])) && c >= 32 => (SRAconst x [31])\n-(SLL x (MOVWconst [c])) => (SLLconst x [c])\n-(SRL x (MOVWconst [...
2020-11-14T12:33:32
huggingface/transformers
b8e5cd5396f7c0cc2d5e10be6696ea38742abf51
1c7ebf1d6eaf0ed0fd4101fd6eb7e64601429cfe
Refactor: Removed un-necessary `object` base class (#32230) * Refactored to remove un-necessary object base class. * small fix.
[ { "path": "examples/research_projects/bertabs/modeling_bertabs.py", "patch": "@@ -557,7 +557,7 @@ def unshape(x):\n return context\n \n \n-class DecoderState(object):\n+class DecoderState:\n \"\"\"Interface for grouping together the current state of a recurrent\n decoder. In the simplest...
2024-07-26T08:33:02
nodejs/node
d8c896cac5b3e31cbffb3da2c657dd59c83fbfeb
eae0c05697121aaae4b097f34d30a259c8591af1
test: use common.expectsError in tests PR-URL: https://github.com/nodejs/node/pull/17484 Reviewed-By: Ruben Bridgewater <ruben@bridgewater.de> Reviewed-By: Colin Ihrig <cjihrig@gmail.com> Reviewed-By: Anatoli Papirovski <apapirovski@mac.com> Reviewed-By: James M Snell <jasnell@gmail.com> Reviewed-By: Luigi Pinca <luig...
[ { "path": "test/parallel/test-buffer-slow.js", "patch": "@@ -58,13 +58,13 @@ const bufferMaxSizeMsg = common.expectsError({\n assert.throws(function() {\n SlowBuffer(Infinity);\n }, bufferMaxSizeMsg);\n-assert.throws(function() {\n+common.expectsError(function() {\n SlowBuffer(-1);\n-}, common.expectsEr...
2017-12-06T08:37:52
ggml-org/llama.cpp
2a619f6fbce255924e8c08a2b82ff96e1d65c2e6
edd4d9bca5dcb9be1ce66e618157d8787a0ea16f
ggml: Vulkan build, Linux -- output error string for errno on fork failure (#20868) (#20904)
[ { "path": "ggml/src/ggml-vulkan/vulkan-shaders/vulkan-shaders-gen.cpp", "patch": "@@ -137,6 +137,7 @@ void execute_command(std::vector<std::string>& command, std::string& stdout_str,\n \n pid_t pid = fork();\n if (pid < 0) {\n+ std::cerr << strerror(errno) << \"\\n\";\n throw std::run...
2026-04-07T11:54:55
rust-lang/rust
256c27e748b477c1ec1c59bdc23afade5eec84aa
281af35cc33e315bd6b52c8a0578108ece75e265
Move methods from `Map` to `TyCtxt`, part 4. Continuing the work from #137350. Removes the unused methods: `expect_variant`, `expect_field`, `expect_foreign_item`. Every method gains a `hir_` prefix.
[ { "path": "compiler/rustc_borrowck/src/diagnostics/mod.rs", "patch": "@@ -505,7 +505,7 @@ impl<'infcx, 'tcx> MirBorrowckCtxt<'_, 'infcx, 'tcx> {\n let var_id =\n self.infcx.tcx.closure_captures(def_id)[field.index()].get_root_variable();\n \n- S...
2025-02-21T07:33:05
golang/go
92c732e901a732855f4b813e6676264421eceae9
782cf560db4c919790fdb476d1bbe18e5ddf5ffd
cmd/compile: fix load of interface{}-typed OpIData in expand_calls In certain cases, the declkared type of an OpIData is interface{}. This was not expected (since interface{} is a pair, right?) and thus caused a crash. What is intended is that these be treated as a byteptr, so do that instead (this is what happens in...
[ { "path": "src/cmd/compile/internal/amd64/ssa.go", "patch": "@@ -76,7 +76,7 @@ func storeByType(t *types.Type) obj.As {\n \t\t\treturn x86.AMOVQ\n \t\t}\n \t}\n-\tpanic(\"bad store type\")\n+\tpanic(fmt.Sprintf(\"bad store type %v\", t))\n }\n \n // moveByType returns the reg->reg move instruction of the gi...
2020-11-13T21:54:48
ollama/ollama
89c79bec8cf7a7b04f761fcc5306d2edf47a4164
c7b77004e35ac86bec8163f63052a5f44122b7d1
Add ModifiedAt Field to /api/show (#5033) * Add Mod Time to Show * Error Handling
[ { "path": "api/types.go", "patch": "@@ -238,6 +238,7 @@ type ShowResponse struct {\n \tSystem string `json:\"system,omitempty\"`\n \tDetails ModelDetails `json:\"details,omitempty\"`\n \tMessages []Message `json:\"messages,omitempty\"`\n+\tModifiedAt time.Time `json:\"modified_at,omitem...
2024-06-16T03:53:56
huggingface/transformers
1c7ebf1d6eaf0ed0fd4101fd6eb7e64601429cfe
c46edfb8230bcc3152e8338742dc4822289acb3d
don't log base model architecture in wandb if log model is false (#32143) * don't log base model architecture in wandb is log model is false * Update src/transformers/integrations/integration_utils.py Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github.com> * convert log model setting into ...
[ { "path": "src/transformers/integrations/integration_utils.py", "patch": "@@ -26,6 +26,7 @@\n import sys\n import tempfile\n from dataclasses import asdict, fields\n+from enum import Enum\n from pathlib import Path\n from typing import TYPE_CHECKING, Any, Dict, Literal, Optional, Union\n \n@@ -726,6 +727,35...
2024-07-26T07:38:59
ggml-org/llama.cpp
71a81f6fcc2c7e4bf17c3c2484c9498358d173b2
ecce0087da749a5d7b0314aae9f1f136eee8b834
ggml-cuda : fix CDNA2 compute capability constant for gfx90a (MI210) (#21519) GGML_CUDA_CC_CDNA2 was set to 0x910 Fix by setting the constant to 0x90a to match the actual gfx90a ISA.
[ { "path": "ggml/src/ggml-cuda/common.cuh", "patch": "@@ -65,7 +65,7 @@\n #define GGML_CUDA_CC_VEGA (GGML_CUDA_CC_OFFSET_AMD + 0x900) // Vega56/64, minimum for fp16 dual issue\n #define GGML_CUDA_CC_VEGA20 (GGML_CUDA_CC_OFFSET_AMD + 0x906) // MI50/Radeon VII, minimum for dp4a\n #define GGML_CUDA_...
2026-04-07T10:18:55
nodejs/node
eae0c05697121aaae4b097f34d30a259c8591af1
20d6b8c6c01c09a51f503f7f8a0fee494be6f742
test: replace assert.throws w/ common.expectsError PR-URL: https://github.com/nodejs/node/pull/17498 Reviewed-By: Anatoli Papirovski <apapirovski@mac.com> Reviewed-By: Jon Moss <me@jonathanmoss.me>
[ { "path": "test/parallel/test-http2-compat-serverresponse-headers.js", "patch": "@@ -72,28 +72,28 @@ server.listen(0, common.mustCall(function() {\n ':path',\n ':authority',\n ':scheme'\n- ].forEach((header) => assert.throws(\n+ ].forEach((header) => common.expectsError(\n () =...
2017-12-06T16:46:44
huggingface/transformers
c46edfb8230bcc3152e8338742dc4822289acb3d
fad15fba78e4603cd20695757ad899a6687485f9
Resize embeds with DeepSpeed (#32214) * fix resize when deepspeed * deepsped uses new embeds * we needed this
[ { "path": "src/transformers/modeling_utils.py", "patch": "@@ -1980,12 +1980,22 @@ def resize_token_embeddings(\n if new_num_tokens is None and pad_to_multiple_of is None:\n return model_embeds\n \n+ # Since we are basically resuing the same old embeddings with new weight values, g...
2024-07-26T05:52:06
golang/go
782cf560db4c919790fdb476d1bbe18e5ddf5ffd
4f63e0a1f88695eec9fc3116d6833b447bcd94a7
cmd/go: permit CGO_LDFLAGS to appear in //go:ldflag Fixes #42565 Change-Id: If7cf39905d124dbd54dfac6a53ee38270498efed Reviewed-on: https://go-review.googlesource.com/c/go/+/269818 Trust: Ian Lance Taylor <iant@golang.org> Run-TryBot: Ian Lance Taylor <iant@golang.org> TryBot-Result: Go Bot <gobot@golang.org> Reviewed...
[ { "path": "src/cmd/go/internal/work/exec.go", "patch": "@@ -2883,6 +2883,21 @@ func (b *Builder) cgo(a *Action, cgoExe, objdir string, pcCFLAGS, pcLDFLAGS, cgo\n \t\t\t\tidx = bytes.Index(src, []byte(cgoLdflag))\n \t\t\t}\n \t\t}\n+\n+\t\t// We expect to find the contents of cgoLDFLAGS in flags.\n+\t\tif le...
2020-11-13T19:05:37
ggml-org/llama.cpp
ecce0087da749a5d7b0314aae9f1f136eee8b834
d1f82e382ddf67f63ee0fd28ae613456cbb23481
fix: Detect streaming state in reasoning content blocks (#21549)
[ { "path": "tools/server/public/index.html", "patch": "@@ -18,7 +18,7 @@\n \t\t<div style=\"display: contents\">\n \t\t\t<script>\n \t\t\t\t{\n-\t\t\t\t\t__sveltekit_1y361v9 = {\n+\t\t\t\t\t__sveltekit_1wqaxod = {\n \t\t\t\t\t\tbase: new URL('.', location).pathname.slice(0, -1)\n \t\t\t\t\t};\n ", "addit...
2026-04-07T10:04:41
nodejs/node
8997026d18cfd54a95c4de6dfdd86bef32f337ff
25385ef26a768817da23746ad00ccf046f8790eb
test: fix flaky async-hooks/test-graph.signal Make sure event loop remains open long enough for signal to be received. PR-URL: https://github.com/nodejs/node/pull/17509 Fixes: https://github.com/nodejs/node/issues/14568 Reviewed-By: Colin Ihrig <cjihrig@gmail.com> Reviewed-By: Richard Lau <riclau@uk.ibm.com> Reviewed...
[ { "path": "test/async-hooks/test-graph.signal.js", "patch": "@@ -3,16 +3,16 @@\n const common = require('../common');\n if (common.isWindows) {\n common.skip('no signals on Windows');\n- return;\n }\n \n const initHooks = require('./init-hooks');\n const verifyGraph = require('./verify-graph');\n-const e...
2017-12-06T22:57:13
huggingface/transformers
4ab33c2d81866d4dd2f29df07f1a35491acbb39b
9d6c0641c4a3c2c5ecf4d49d7609edd5b745d9bc
Generation: stop at `eos` for assisted decoding (#31301) * fix * move changes to prompt lookup * add test * set eos in assistant model * style * fix flakiness * changes for new `main` * Update tests/generation/test_utils.py Co-authored-by: amyeroberts <22614925+amyeroberts@users.noreply.github....
[ { "path": "src/transformers/generation/candidate_generator.py", "patch": "@@ -108,6 +108,9 @@ def __init__(\n self.assistant_model = assistant_model\n self.num_assistant_tokens = assistant_model.generation_config.num_assistant_tokens\n \n+ # Set eos in assistant same as in target mode...
2024-07-26T05:16:06
ollama/ollama
225f0d121948f4d9ca8700d66c3514f233feaffd
532db583113533a0f512568c95eb448628873415
gpu: Fix build warning Signed-off-by: Lei Jitang <leijitang@outlook.com>
[ { "path": "gpu/gpu_info_oneapi.c", "patch": "@@ -160,7 +160,7 @@ void oneapi_check_vram(oneapi_handle_t h, int driver, int device,\n return;\n }\n \n- snprintf(&resp->gpu_name[0], GPU_NAME_LEN, props.modelName);\n+ snprintf(&resp->gpu_name[0], GPU_NAME_LEN, \"%s\", props.modelName);\n \n // TODO t...
2024-06-15T06:26:23
rust-lang/rust
faa5b3f7de0db76e41bcf5c54c401d7651e69c58
3ea711f17e3946ac3f4df11691584e2c56b4b0cf
Fix false-positive in `expr_or_init` and in the `invalid_from_utf8` lint
[ { "path": "compiler/rustc_lint/src/context.rs", "patch": "@@ -6,6 +6,7 @@\n use std::cell::Cell;\n use std::slice;\n \n+use rustc_ast::BindingMode;\n use rustc_data_structures::fx::FxIndexMap;\n use rustc_data_structures::sync;\n use rustc_data_structures::unord::UnordMap;\n@@ -14,6 +15,7 @@ use rustc_featu...
2025-03-09T17:49:59
ggml-org/llama.cpp
d1f82e382ddf67f63ee0fd28ae613456cbb23481
0988accf824e817d0e3db55c05f5f2954fbfe505
Fix rtl text rendering (#21382) * Fix Arabic RTL text rendering in web UI - Add dir='auto' attributes to markdown containers and blocks - Implement post-processing to add dir='auto' to all text elements - Replace directional CSS properties with logical properties for proper RTL list alignment - Ensure bidirectional t...
[ { "path": "tools/server/public/index.html", "patch": "@@ -18,7 +18,7 @@\n \t\t<div style=\"display: contents\">\n \t\t\t<script>\n \t\t\t\t{\n-\t\t\t\t\t__sveltekit_1trm5n9 = {\n+\t\t\t\t\t__sveltekit_1y361v9 = {\n \t\t\t\t\t\tbase: new URL('.', location).pathname.slice(0, -1)\n \t\t\t\t\t};\n ", "addit...
2026-04-07T09:37:20
nodejs/node
acd427713422138a92f74a4c3c2a1f6d8ecf9778
20a8e83e44a6555a40c9d5f367a9a7be8fed0374
test: replace assert.throws w/ common.expectsError PR-URL: https://github.com/nodejs/node/pull/17497 Reviewed-By: Ruben Bridgewater <ruben@bridgewater.de> Reviewed-By: Jon Moss <me@jonathanmoss.me> Reviewed-By: Luigi Pinca <luigipinca@gmail.com>
[ { "path": "test/parallel/test-dns.js", "patch": "@@ -68,16 +68,16 @@ const goog = [\n ];\n assert.doesNotThrow(() => dns.setServers(goog));\n assert.deepStrictEqual(dns.getServers(), goog);\n-assert.throws(() => dns.setServers(['foobar']), common.expectsError({\n+common.expectsError(() => dns.setServers(['f...
2017-12-06T15:38:06
huggingface/transformers
9d6c0641c4a3c2c5ecf4d49d7609edd5b745d9bc
3a83ec48a63a8298c8193be48cf00785674bfb70
Fix code snippet for Grounding DINO (#32229) Fix code snippet for grounding-dino
[ { "path": "docs/source/en/model_doc/grounding-dino.md", "patch": "@@ -41,33 +41,40 @@ The original code can be found [here](https://github.com/IDEA-Research/Grounding\n Here's how to use the model for zero-shot object detection:\n \n ```python\n-import requests\n-\n-import torch\n-from PIL import Image\n-fr...
2024-07-25T18:20:47
golang/go
86954d5246339231dc0c45f5547c37a1c3650264
f423d616b15302730c1b737a3b22afca315a7fbe
cmd/compile: mark plugin-exported types as used in interface Plugin exports symbols as interfaces. Mark their types as used in interfaces, so their methods will be kept alive by the linker. Fixes #42579. Change-Id: If1b5aacc21510c20c25f88bb131bca61db6f1d56 Reviewed-on: https://go-review.googlesource.com/c/go/+/26981...
[ { "path": "misc/cgo/testplugin/plugin_test.go", "patch": "@@ -196,3 +196,17 @@ func TestIssue25756(t *testing.T) {\n \t\t})\n \t}\n }\n+\n+func TestMethod(t *testing.T) {\n+\t// Exported symbol's method must be live.\n+\tgoCmd(t, \"build\", \"-buildmode=plugin\", \"-o\", \"plugin.so\", \"./method/plugin.go\...
2020-11-13T19:42:45
rust-lang/rust
53f488aa4b5be28c3cd350cde133edc0efe177ff
9fb94b32df38073bf63d009df77ed10cb1c989d0
Simulate OOM for the `try_oom_error` test We can create the expected error manually, rather than trying to produce a real one, so the error conversion test can run on all targets. Before, it was only running on 64-bit and not miri. In Fedora, we also found that s390x was not getting the expected error, "successfully"...
[ { "path": "library/std/src/io/tests.rs", "patch": "@@ -811,13 +811,17 @@ fn read_to_end_error() {\n }\n \n #[test]\n-// Miri does not support signalling OOM\n-#[cfg_attr(miri, ignore)]\n-// 64-bit only to be sure the allocator will fail fast on an impossible to satisfy size\n-#[cfg(target_pointer_width = \"...
2025-03-11T20:27:15
ggml-org/llama.cpp
0988accf824e817d0e3db55c05f5f2954fbfe505
0033f53a072af953b457c9fd2314e6e28bd11cc7
[SYCL] Add Q8_0 reorder optimization (~3x tg speedup on Intel Arc) (#21527) Extend the existing reorder optimization to Q8_0. The reorder separates scale factors from weight data for coalesced memory access -- was implemented for Q4_0/Q4_K/Q6_K but Q8_0 was missing. On Arc Pro B70 (Xe2), Q8_0 tg goes from 4.88 to 15....
[ { "path": "ggml/src/ggml-sycl/dequantize.hpp", "patch": "@@ -143,6 +143,22 @@ static __dpct_inline__ void dequantize_q5_1(const void *vx, const int64_t ib,\n #endif // GGML_SYCL_F16\n }\n \n+static __dpct_inline__ void dequantize_q8_0_reorder(const void *d_ptr, const int64_t ib, const void *qs,\n+ ...
2026-04-07T08:12:49
nodejs/node
20a8e83e44a6555a40c9d5f367a9a7be8fed0374
f81bb7b527bb58ee68ddd26c9eee56a133547c01
test: replace assert.throws w/ common.expectsError PR-URL: https://github.com/nodejs/node/pull/17494 Reviewed-By: Anatoli Papirovski <apapirovski@mac.com> Reviewed-By: James M Snell <jasnell@gmail.com> Reviewed-By: Colin Ihrig <cjihrig@gmail.com> Reviewed-By: Ruben Bridgewater <ruben@bridgewater.de> Reviewed-By: Jon M...
[ { "path": "test/parallel/test-dgram-createSocket-type.js", "patch": "@@ -23,13 +23,13 @@ const errMessage = /^Bad socket type specified\\. Valid types are: udp4, udp6$/;\n \n // Error must be thrown with invalid types\n invalidTypes.forEach((invalidType) => {\n- assert.throws(() => {\n+ common.expectsErro...
2017-12-06T14:32:42
huggingface/transformers
df6eee9201e4ba2b80cea021a18e95ada26ca2cc
de2318894e4f971ea2273c653a702dc93db2bd6a
Follow up for #31973 (#32025) * fix * [test_all] trigger full CI --------- Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>
[ { "path": "tests/generation/test_configuration_utils.py", "patch": "@@ -18,10 +18,10 @@\n import tempfile\n import unittest\n import warnings\n+from pathlib import Path\n \n from huggingface_hub import HfFolder, delete_repo\n from parameterized import parameterized\n-from requests.exceptions import HTTPErro...
2024-07-25T14:12:23
golang/go
f423d616b15302730c1b737a3b22afca315a7fbe
35455fff0ebb7dd1b8e698f245a823ef8c711ac9
cmd/cgo: fix initialization of empty argument types CL 258938 changed the way C to Go calls work such that they now construct a C struct on the C side for the arguments and space for the results. Any pointers in the result space must be zeroed, so we just zero the whole struct. However, C makes it surprisingly hard t...
[ { "path": "misc/cgo/test/issue42495.go", "patch": "@@ -0,0 +1,15 @@\n+// Copyright 2020 The Go Authors. All rights reserved.\n+// Use of this source code is governed by a BSD-style\n+// license that can be found in the LICENSE file.\n+\n+package cgotest\n+\n+// typedef struct { } T42495A;\n+// typedef struc...
2020-11-11T20:33:39
ollama/ollama
da3bf2335483fac7cbfff72a1b80e40988f4f35a
45cacbaf0568a4d38d74ebdd0957fe01bd06719d
Workaround gfx900 SDMA bugs Implement support for GPU env var workarounds, and leverage this for the Vega RX 56 which needs HSA_ENABLE_SDMA=0 set to work properly
[ { "path": "gpu/amd_linux.go", "patch": "@@ -332,6 +332,11 @@ func AMDGetGPUInfo() []RocmGPUInfo {\n \t\t\tslog.Info(\"skipping rocm gfx compatibility check\", \"HSA_OVERRIDE_GFX_VERSION\", gfxOverride)\n \t\t}\n \n+\t\t// Check for env var workarounds\n+\t\tif name == \"1002:687f\" { // Vega RX 56\n+\t\t\tg...
2024-05-31T23:15:21
ggml-org/llama.cpp
0033f53a072af953b457c9fd2314e6e28bd11cc7
d0a6dfeb28a09831d904fc4d910ddb740da82834
docs: fix typo in build.md (emdawbwebgpu -> emdawnwebgpu) (#21518)
[ { "path": "docs/build.md", "patch": "@@ -741,7 +741,7 @@ cmake --build build --config Release\n \n WebGPU allows cross-platform access to the GPU from supported browsers. We utilize [Emscripten](https://emscripten.org/) to compile ggml's WebGPU backend to WebAssembly. Emscripten does not officially support ...
2026-04-07T04:37:26
huggingface/transformers
de2318894e4f971ea2273c653a702dc93db2bd6a
9b9a54e61bf8749588178b37c23d77b90679fd10
[warnings] fix E721 warnings (#32223) fix E721 warnings
[ { "path": "src/transformers/generation/candidate_generator.py", "patch": "@@ -162,7 +162,7 @@ def __init__(\n self.generation_config.min_length = 0\n self.generation_config.min_new_tokens = None\n for processor in self.logits_processor:\n- if type(processor) == MinLengthLo...
2024-07-25T13:12:23
nodejs/node
2ac4718ac4e61545872140f3b8eca496f80b2e8a
6fa9cae83f859cbe73a98cfc2ea18ad0f1e2b87e
2017-12-08, Version 8.9.3 'Carbon' (LTS) This is a security release. All Node.js users should consult the security release summary at https://nodejs.org/en/blog/vulnerability/december-2017-security-releases/ for details on patched vulnerabilities. Fixes for the following CVEs are included in this release: * CVE-2017...
[ { "path": "CHANGELOG.md", "patch": "@@ -34,7 +34,8 @@ release.\n <a href=\"doc/changelogs/CHANGELOG_V9.md#9.0.0\">9.0.0</a><br/>\n </td>\n <td valign=\"top\">\n-<b><a href=\"doc/changelogs/CHANGELOG_V8.md#8.9.2\">8.9.2</a></b><br/>\n+<b><a href=\"doc/changelogs/CHANGELOG_V8.md#8.9.3\">8.9.3</a></b><...
2017-12-07T20:07:03
ggml-org/llama.cpp
2e1f0a889e19a3922db57452268f4574c35c36e5
506200cf8b5c8419ce97d16dc8c50f4634e21ebe
ggml: add Q1_0 1-bit quantization support (CPU) (#21273) * ggml: add Q1_0 and Q1_0_g128 1-bit quantization support (CPU) * add generic fallback for x86 * remove Q1_0 (group size 32) * rename Q1_0_g128 => Q1_0 * fix Q1_0 LlamaFileType Enum * Fix trailing spaces; add generic fallback for othre backends * Apply sug...
[ { "path": "ggml/include/ggml.h", "patch": "@@ -428,7 +428,8 @@ extern \"C\" {\n // GGML_TYPE_IQ4_NL_8_8 = 38,\n GGML_TYPE_MXFP4 = 39, // MXFP4 (1 block)\n GGML_TYPE_NVFP4 = 40, // NVFP4 (4 blocks, E4M3 scale)\n- GGML_TYPE_COUNT = 41,\n+ GGML_TYPE_Q1_0 = 41,\n+ ...
2026-04-06T18:55:21
golang/go
35455fff0ebb7dd1b8e698f245a823ef8c711ac9
31f71506d7026595be76713af25197a8c0022ac8
runtime: swap the order of raceacquire() and racerelease() In chansend() and chanrecv() of chan.go, the order of calls to raceacquire() and racerelease() was swapped, which meant that the code was not following the memory model "by the letter of the law." Similar for bufrecv and bufsend in select.go The memory model ...
[ { "path": "src/runtime/chan.go", "patch": "@@ -215,8 +215,7 @@ func chansend(c *hchan, ep unsafe.Pointer, block bool, callerpc uintptr) bool {\n \t\t// Space is available in the channel buffer. Enqueue the element to send.\n \t\tqp := chanbuf(c, c.sendx)\n \t\tif raceenabled {\n-\t\t\traceacquire(qp)\n-\t\t...
2020-11-12T16:25:47
ollama/ollama
206797bda4685288de9a775b1536b9cbde0a7246
43ed358f9a894c92a24d06346ad81dc76aa52cfb
Fix concurrency integration test to work locally This worked remotely but wound up trying to spawn multiple servers locally which doesn't work
[ { "path": "integration/concurrency_test.go", "patch": "@@ -45,10 +45,18 @@ func TestMultiModelConcurrency(t *testing.T) {\n \twg.Add(len(req))\n \tctx, cancel := context.WithTimeout(context.Background(), time.Second*120)\n \tdefer cancel()\n+\n+\tclient, _, cleanup := InitServerConnection(ctx, t)\n+\tdefer ...
2024-05-23T20:12:14
huggingface/transformers
5658e749adbaaf883caec003cecae8ce0a4261a6
85a1269e19af022e04bc2aad82572cd5a9e8cdd9
[whisper] fix short-form output type (#32178) * [whisper] fix short-form output type * add test * make style * update long-form tests * fixes * last fix * finalise test
[ { "path": "src/transformers/models/whisper/generation_whisper.py", "patch": "@@ -498,7 +498,7 @@ def generate(\n \n # 3. Make sure generation config is correctly set\n # Make sure the generation config is correctly set depending on whether timestamps are to be returned or not\n- self....
2024-07-25T08:58:02
nodejs/node
6fa9cae83f859cbe73a98cfc2ea18ad0f1e2b87e
e832743f72229053f6c6e0dcefb41170e60cf301
2017-12-08, Version 6.12.2 'Boron' (LTS) This is a security release. All Node.js users should consult the security release summary at https://nodejs.org/en/blog/vulnerability/december-2017-security-releases/ for details on patched vulnerabilities. Fixes for the following CVEs are included in this release: * CVE-2017...
[ { "path": "CHANGELOG.md", "patch": "@@ -54,7 +54,8 @@ release.\n <a href=\"doc/changelogs/CHANGELOG_V8.md#8.0.0\">8.0.0</a><br/>\n </td>\n <td valign=\"top\">\n-<b><a href=\"doc/changelogs/CHANGELOG_V6.md#6.12.1\">6.12.1</a></b><br/>\n+<b><a href=\"doc/changelogs/CHANGELOG_V6.md#6.12.2\">6.12.2</a></b...
2017-12-07T20:14:20
ggml-org/llama.cpp
506200cf8b5c8419ce97d16dc8c50f4634e21ebe
15f786e6581598638840276948a7e6183fc96a83
cli: fix stripping of \n in multiline input (#21485) * llama-cli: fix stripping of \n in multiline input * Change & string to string_view * Apply suggestions from code review Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com> * Fix EditorConfig linter error --------- Co-authored-by: Sigbjørn Skjæret ...
[ { "path": "common/console.cpp", "patch": "@@ -700,13 +700,13 @@ namespace console {\n std::vector<std::string> entries;\n size_t viewing_idx = SIZE_MAX;\n std::string backup_line; // current line before viewing history\n- void add(const std::string & line) {\n+ void add...
2026-04-06T18:54:06
huggingface/transformers
85a1269e19af022e04bc2aad82572cd5a9e8cdd9
edd68f4ed8db241bd3e9dc6c4ed96d471f243c9a
fix: Replaced deprecated `unittest method` with the correct one (#32198) Replaced deprecated unittest method with the correct one.
[ { "path": "tests/models/llava/test_processor_llava.py", "patch": "@@ -44,4 +44,4 @@ def test_chat_template(self):\n ]\n \n formatted_prompt = processor.apply_chat_template(messages, add_generation_prompt=True)\n- self.assertEquals(expected_prompt, formatted_prompt)\n+ self.asse...
2024-07-24T17:00:21
ollama/ollama
fb9cdfa72335993e454af1c0eee2235f2a7e88a4
efac488675bb4118b6dab2587e481e8640102fad
Fix server.cpp for the new cuda build macros
[ { "path": "llm/ext_server/server.cpp", "patch": "@@ -2335,9 +2335,9 @@ static void server_params_parse(int argc, char **argv, server_params &sparams, g\n invalid_param = true;\n break;\n }\n-#ifndef GGML_USE_CUBLAS\n- fprintf(stderr, \"warning: llama.cp...
2024-05-18T23:02:13
ggml-org/llama.cpp
15f786e6581598638840276948a7e6183fc96a83
94ca829b6001019622c0f67fcd48e9ec6bd7dce8
[CUDA ] Write an optimized flash_attn_stream_k_fixup kernel (#21159) * Write an optimized flash_attn_stream_k_fixup kernel Write a specialized and more optimized kernel for cases where nblocks_stream_k is multiple of ntiles_dst. Make nblocks_stream_k to multiple of ntiles_dst if nblocks_stream_k > 2 * ntiles_dst * U...
[ { "path": "ggml/src/ggml-cuda/fattn-common.cuh", "patch": "@@ -676,9 +676,96 @@ static __global__ void flash_attn_mask_to_KV_max(\n \n template<int D, int ncols1, int ncols2> // D == head size\n __launch_bounds__(D, 1)\n-static __global__ void flash_attn_stream_k_fixup(\n- float * __restrict__ dst, c...
2026-04-06T18:34:29
nodejs/node
e832743f72229053f6c6e0dcefb41170e60cf301
07549c6a91f8901b69bfc430b1a552deca9763b8
2017-12-08, Version 4.8.7 'Argon' (LTS) This is a security release. All Node.js users should consult the security release summary at https://nodejs.org/en/blog/vulnerability/december-2017-security-releases/ for details on patched vulnerabilities. Fixes for the following CVEs are included in this release: * CVE-2017-...
[ { "path": "CHANGELOG.md", "patch": "@@ -87,7 +87,8 @@ release.\n <a href=\"doc/changelogs/CHANGELOG_V6.md#6.0.0\">6.0.0</a><br/>\n </td>\n <td valign=\"top\">\n-<b><a href=\"doc/changelogs/CHANGELOG_V4.md#4.8.6\">4.8.6</a></b><br/>\n+<b><a href=\"doc/changelogs/CHANGELOG_V4.md#4.8.7\">4.8.7</a></b><br/>...
2017-12-07T20:24:08
huggingface/transformers
edd68f4ed8db241bd3e9dc6c4ed96d471f243c9a
1c122a46dc3c4448901f8d2f3018d9d58b846ba5
:rotating_light: No more default chat templates (#31733) * No more default chat templates * Add the template to the GPT-SW3 tests since it's not available by default now * Fix GPT2 test * Fix Bloom test * Fix Bloom test * Remove default templates again
[ { "path": "docs/source/en/chat_templating.md", "patch": "@@ -580,7 +580,7 @@ default template for that model class is used instead. Let's take a look at the\n >>> from transformers import AutoTokenizer\n >>> tokenizer = AutoTokenizer.from_pretrained(\"facebook/blenderbot-400M-distill\")\n \n->>> tokenizer.d...
2024-07-24T16:36:32
golang/go
062e0e5ce6df339dc26732438ad771f73dbf2292
1e1fa5903b760c6714ba17e50bf850b01f49135c
cmd/go, cmd/cgo: don't let bogus symbol set cgo_ldflag A hand-edited object file can have a symbol name that uses newline and other normally invalid characters. The cgo tool will generate Go files containing symbol names, unquoted. That can permit those symbol names to inject Go code into a cgo-generated file. If that...
[ { "path": "misc/cgo/errors/badsym_test.go", "patch": "@@ -0,0 +1,216 @@\n+// Copyright 2020 The Go Authors. All rights reserved.\n+// Use of this source code is governed by a BSD-style\n+// license that can be found in the LICENSE file.\n+\n+package errorstest\n+\n+import (\n+\t\"bytes\"\n+\t\"io/ioutil\"\n...
2020-11-06T17:38:38
ggml-org/llama.cpp
941146b3f1ebcd54f125f1f80598f29231155989
482d862bcbf813f5a8393ac05a2c31c647fc2984
convert : fix block_ff_dim retrieval for lfm2 (#21508)
[ { "path": "convert_hf_to_gguf.py", "patch": "@@ -11818,10 +11818,8 @@ class LFM2Model(TextModel):\n model_arch = gguf.MODEL_ARCH.LFM2\n \n def _add_feed_forward_length(self):\n- ff_dim = self.hparams[\"block_ff_dim\"]\n-\n+ ff_dim = self.find_hparam([\"block_ff_dim\", \"intermediate_si...
2026-04-06T12:05:18
huggingface/transformers
af0e4b7b37b2d7eefe7531cf5201a5d6bae85525
1392a6867f40a55dfabaf306745c67627598b1af
Fix float8_e4m3fn in modeling_utils (#32193) * Fix float8_e4m3fn in modeling_utils * style * fix * comment
[ { "path": "src/transformers/modeling_utils.py", "patch": "@@ -855,6 +855,8 @@ def _load_state_dict_into_meta_model(\n for old_key, new_key in zip(old_keys, new_keys):\n state_dict[new_key] = state_dict.pop(old_key)\n \n+ is_torch_e4m3fn_available = hasattr(torch, \"float8_e4m3fn\")\n+\n f...
2024-07-24T15:14:05
nodejs/node
51c6737d3c733718fd575f79bcffbba62a4b57fc
abc280172f1cdd44c8dbb8285a67322b9e16a414
src: remove unused variable in node_contextify Currently the following warning is show when building: ../src/node_contextify.cc:642:10: warning: unused variable 'display_errors' [-Wunused-variable] bool display_errors = maybe_display_errors.ToChecked(); This commit removes the unused variable which became unused...
[ { "path": "src/node_contextify.cc", "patch": "@@ -629,7 +629,6 @@ class ContextifyScript : public BaseObject {\n MaybeLocal<String> filename = GetFilenameArg(env, options);\n MaybeLocal<Integer> lineOffset = GetLineOffsetArg(env, options);\n MaybeLocal<Integer> columnOffset = GetColumnOffsetArg(...
2017-12-06T13:10:35
ollama/ollama
d528e1af75e5b253b21ec90fa9d5f33737fd4909
cd234ce22c85bf34dc50b05c93c4dab513ae8f99
fix utf16 for multibyte runes
[ { "path": "parser/parser.go", "patch": "@@ -8,7 +8,9 @@ import (\n \t\"io\"\n \t\"strconv\"\n \t\"strings\"\n-\t\"unicode\"\n+\n+\t\"golang.org/x/text/encoding/unicode\"\n+\t\"golang.org/x/text/transform\"\n )\n \n type File struct {\n@@ -69,14 +71,11 @@ func ParseFile(r io.Reader) (*File, error) {\n \tvar ...
2024-06-13T18:39:01
ggml-org/llama.cpp
58190cc84d846d8575ba26e8486bc29d9fd8ad55
af76639f728c69c74c873cc45f038aaa17afd09e
llama : correct platform-independent loading of BOOL metadata (#21428) * model-loader : fix GGUF bool array conversion * model-loader : fix remaining GGUF bool pointer uses
[ { "path": "src/llama-impl.cpp", "patch": "@@ -128,7 +128,7 @@ static std::string gguf_data_to_str(enum gguf_type type, const void * data, int\n case GGUF_TYPE_INT64: return std::to_string(((const int64_t *)data)[i]);\n case GGUF_TYPE_FLOAT32: return std::to_string(((const float *)data)...
2026-04-05T23:40:38
golang/go
1e1fa5903b760c6714ba17e50bf850b01f49135c
b34b0aaf69349f060d3b03a06f520848964cb7eb
math/big: fix shift for recursive division The previous s value could cause a crash for certain inputs. Will check in tests and documentation improvements later. Thanks to the Go Ethereum team and the OSS-Fuzz project for reporting this. Thanks to Rémy Oudompheng and Robert Griesemer for their help developing and va...
[ { "path": "src/math/big/nat.go", "patch": "@@ -929,7 +929,7 @@ func (z nat) divRecursiveStep(u, v nat, depth int, tmp *nat, temps []*nat) {\n \n \t// Now u < (v<<B), compute lower bits in the same way.\n \t// Choose shift = B-1 again.\n-\ts := B\n+\ts := B - 1\n \tqhat := *temps[depth]\n \tqhat.clear()\n \t...
2020-11-10T20:54:12
huggingface/transformers
1392a6867f40a55dfabaf306745c67627598b1af
8d2534c4d0ab94a97a72d2ce6bb9ccd201abadb3
Fix resize embedding with Deepspeed (#32192) fix resize when deepspeed
[ { "path": "src/transformers/modeling_utils.py", "patch": "@@ -2131,13 +2131,23 @@ def _get_resized_embeddings(\n # Replace weights in old_embeddings and return to maintain the same embedding type.\n # This ensures correct functionality when a Custom Embedding class is passed as input.\n ...
2024-07-24T14:26:20
nodejs/node
09e9e2bd82d5b82747100ab9163ebadba4cb3972
e55b7d62863b7eae82e4afccf0b874e0f5ebba16
test: fix flaky test-benchmark-querystring Allow zero iterations for short benchmark in test. PR-URL: https://github.com/nodejs/node/pull/17517 Reviewed-By: Anatoli Papirovski <apapirovski@mac.com> Reviewed-By: Colin Ihrig <cjihrig@gmail.com> Reviewed-By: Evan Lucas <evanlucas@me.com> Reviewed-By: Refael Ackermann <r...
[ { "path": "test/parallel/test-benchmark-querystring.js", "patch": "@@ -4,8 +4,9 @@ require('../common');\n \n const runBenchmark = require('../common/benchmark');\n \n-runBenchmark('querystring', [\n- 'n=1',\n- 'input=\"there is nothing to unescape here\"',\n- 'type=noencode'\n-]);\n+runBenchmark('querys...
2017-12-07T03:39:07
ollama/ollama
1fd236d1776445b45af104d259a0d997ef0b1806
e87fc7200d47bd2e5214126802da1d5607b686be
server: remove jwt decoding error (#5027)
[ { "path": "server/images.go", "patch": "@@ -960,7 +960,6 @@ var errUnauthorized = fmt.Errorf(\"unauthorized: access denied\")\n func getTokenSubject(token string) string {\n \tparts := strings.Split(token, \".\")\n \tif len(parts) != 3 {\n-\t\tslog.Error(\"jwt token does not contain 3 parts\")\n \t\treturn ...
2024-06-13T18:21:15
ggml-org/llama.cpp
af76639f728c69c74c873cc45f038aaa17afd09e
761797ffdf2ce3f118e82c663b1ad7d935fbd656
model : add HunyuanOCR support (#21395) * HunyuanOCR: add support for text and vision models - Add HunyuanOCR vision projector (perceiver-based) with Conv2d merge - Add separate HUNYUAN_OCR chat template (content-before-role format) - Handle HunyuanOCR's invalid pad_token_id=-1 in converter - Fix EOS/EOT token IDs fr...
[ { "path": "convert_hf_to_gguf.py", "patch": "@@ -11521,13 +11521,50 @@ def prepare_tensors(self):\n raise ValueError(f\"Unprocessed experts: {experts}\")\n \n \n-@ModelBase.register(\"HunYuanDenseV1ForCausalLM\")\n+@ModelBase.register(\"HunYuanDenseV1ForCausalLM\", \"HunYuanVLForConditionalG...
2026-04-05T21:32:14
golang/go
b34b0aaf69349f060d3b03a06f520848964cb7eb
c167635a6e08cdb1223a6d7b70cffd73c165562e
cmd/go: skip TestScript/build_plugin_non_main on platforms that do not support -buildmode=plugin Fixes #42474 Change-Id: I1550b44b92cd272854e2f17493245a14e3d39f41 GitHub-Last-Rev: 948d01716eff41c25515dfb6135769862977aba1 GitHub-Pull-Request: golang/go#42475 Reviewed-on: https://go-review.googlesource.com/c/go/+/26873...
[ { "path": "src/cmd/go/testdata/script/build_plugin_non_main.txt", "patch": "@@ -1,7 +1,5 @@\n-# Plugins are only supported on linux,cgo (!riscv64) and darwin,cgo.\n-[!linux] [!darwin] skip\n-[linux] [riscv64] skip\n-[!cgo] skip\n+# Plugins are not supported on all platforms.\n+[!buildmode:plugin] skip\n \n ...
2020-11-12T08:12:08
huggingface/transformers
8d2534c4d0ab94a97a72d2ce6bb9ccd201abadb3
e0182f3bd7f4753c1e378e052ceea67898d97359
let's not warn when someone is running a forward (#32176) * let's not warn when someone is running a foward without cache + self.training * more models * fixup
[ { "path": "src/transformers/models/cohere/modeling_cohere.py", "patch": "@@ -769,7 +769,9 @@ def forward(\n \n past_seen_tokens = 0\n return_legacy_cache = False\n- if use_cache and not isinstance(past_key_values, Cache): # kept for BC (non `Cache` `past_key_values` inputs)\n+ ...
2024-07-24T14:06:39
nodejs/node
a413df40b67e1005b6c0c6d9e280bd7dcb53ee11
de06dc92d150c1f999462a18c7594a5eea81225f
doc: fix typo in repl.md PR-URL: https://github.com/nodejs/node/pull/17502 Reviewed-By: Colin Ihrig <cjihrig@gmail.com> Reviewed-By: Richard Lau <riclau@uk.ibm.com> Reviewed-By: Anatoli Papirovski <apapirovski@mac.com> Reviewed-By: Jon Moss <me@jonathanmoss.me> Reviewed-By: Minwoo Jung <minwoo@nodesource.com> Reviewed...
[ { "path": "doc/api/repl.md", "patch": "@@ -418,7 +418,7 @@ changes:\n * `output` {Writable} The Writable stream to which REPL output will be\n written. Defaults to `process.stdout`.\n * `terminal` {boolean} If `true`, specifies that the `output` should be\n- treated as a a TTY terminal, and have ...
2017-12-06T18:18:05
ggml-org/llama.cpp
5d3a4a7da5e3dd42f5922aba2fe21b520e96e830
c08d28d08871715fd68accffaeeb76ddcaede658
server : fix logging of build + system info (#21460) This PR changes the logging that occurs at startup of llama-server. Currently, it is redundant (including CPU information twice) and it is missing the build + commit info.
[ { "path": "tools/server/server.cpp", "patch": "@@ -108,10 +108,8 @@ int main(int argc, char ** argv) {\n llama_backend_init();\n llama_numa_init(params.numa);\n \n- LOG_INF(\"system info: n_threads = %d, n_threads_batch = %d, total_threads = %d\\n\", params.cpuparams.n_threads, params.cpuparams_b...
2026-04-05T14:14:02
huggingface/transformers
5f4ee98a7ade33e1c54fdd6181d04ee7b426b392
8678879f1dc2578cec18232146bf19de97aecaa1
Update qwen2.md (#32108) * Update qwen2.md outdated description * Update qwen2.md amended * Update qwen2.md Update * Update qwen2.md fix wrong version code, now good to go
[ { "path": "docs/source/en/model_doc/qwen2.md", "patch": "@@ -18,7 +18,7 @@ rendered properly in your Markdown viewer.\n \n ## Overview\n \n-Qwen2 is the new model series of large language models from the Qwen team. Previously, we released the Qwen series, including Qwen-72B, Qwen-1.8B, Qwen-VL, Qwen-Audio, ...
2024-07-24T10:54:41
golang/go
9ef65ff137c17bb7975859017abae4e27600c684
4bc5f6f45f6e887f74b22dfdfffd6df2a2f97094
cmd/compile: do not emit an extra debug_line entry for the end of seq addr Uses DW_LNS_advance_pc directly, instead of calling putpclcdelta because the latter will create a new debug_line entry for the end of sequence address. Fixes #42484 Change-Id: Ib6355605cac101b9bf37a3b4961ab0cee678a839 Reviewed-on: https://go-...
[ { "path": "src/cmd/internal/obj/dwarf.go", "patch": "@@ -104,7 +104,8 @@ func (ctxt *Link) generateDebugLinesSymbol(s, lines *LSym) {\n \t// GDB will assign a line number of zero the last row in the line\n \t// table, which we don't want.\n \tlastlen := uint64(s.Size - (lastpc - s.Func().Text.Pc))\n-\tputpc...
2020-11-10T16:02:27
nodejs/node
de06dc92d150c1f999462a18c7594a5eea81225f
86d386a07d3bb7d13c0b174f37cf2bc1d00dee51
doc: fix common typo involving one-time listeners Our docs use both "one time listener" and "one-time listener". The second is more correct. Standardize on that. PR-URL: https://github.com/nodejs/node/pull/17502 Reviewed-By: Colin Ihrig <cjihrig@gmail.com> Reviewed-By: Richard Lau <riclau@uk.ibm.com> Reviewed-By: Ana...
[ { "path": "doc/api/events.md", "patch": "@@ -409,7 +409,7 @@ added: v0.3.0\n * `eventName` {any} The name of the event.\n * `listener` {Function} The callback function\n \n-Adds a **one time** `listener` function for the event named `eventName`. The\n+Adds a **one-time** `listener` function for the event na...
2017-12-06T18:15:58
ggml-org/llama.cpp
661e9acb36768d0d4ddb6f2eb674fbb1be185823
b8635075ffe27b135c49afb9a8b5c434bd42c502
ci: fix vulkan workflow referencing non-existent action (#21442)
[ { "path": ".github/workflows/build-vulkan.yml", "patch": "@@ -72,7 +72,7 @@ jobs:\n \n - name: Setup Vulkan SDK\n if: steps.cache-sdk.outputs.cache-hit != 'true'\n- uses: ./.github/actions/linux-setup-vulkan-llvmpipe\n+ uses: ./.github/actions/linux-setup-vulkan\n with:\n...
2026-04-05T00:59:51
ollama/ollama
c16f8af91150febf705cefe28d48c90979cece30
217f60c3d95edc7d5dcbeb4c3cffb0190c147f92
fix: multiple templates when creating from model multiple templates may appear in a model if a model is created from another model that 1) has an autodetected template and 2) defines a custom template
[ { "path": "server/images.go", "patch": "@@ -28,7 +28,6 @@ import (\n \t\"github.com/ollama/ollama/format\"\n \t\"github.com/ollama/ollama/llm\"\n \t\"github.com/ollama/ollama/parser\"\n-\t\"github.com/ollama/ollama/templates\"\n \t\"github.com/ollama/ollama/types/errtypes\"\n \t\"github.com/ollama/ollama/ty...
2024-06-12T20:30:08
huggingface/transformers
8678879f1dc2578cec18232146bf19de97aecaa1
01be5b48790f113b7d71943b580c842e3e097988
fix: default value reflects the runtime environment variables rather than the ones present at import time. (#32153) * fix: default value reflects the runtime environment variables rather than the ones present at import time. * Fix: Change `deterministic` to None by default; use env var if None
[ { "path": "src/transformers/modeling_flash_attention_utils.py", "patch": "@@ -193,7 +193,7 @@ def _flash_attention_forward(\n sliding_window: Optional[int] = None,\n use_top_left_mask: bool = False,\n softcap: Optional[float] = None,\n- deterministic: bool = os.environ.get(\"FLASH_ATTENTION_D...
2024-07-24T10:38:49
golang/go
4bc5f6f45f6e887f74b22dfdfffd6df2a2f97094
d7974c31d0eb0ef377a8681f6f7306d46854eb1c
cmd/link: put DYLD_INFO at beginning of LINKEDIT segment on darwin Apparently, code signing requires DYLD_INFO tables are at the beginning of the LINKEDIT segment. Put it there. May fix #42507. Change-Id: I1836e0f495719cf75f66d0831fe1544bbe3ff1a8 Reviewed-on: https://go-review.googlesource.com/c/go/+/269377 Trust: C...
[ { "path": "src/cmd/link/internal/ld/macho.go", "patch": "@@ -761,12 +761,12 @@ func asmbMacho(ctxt *Link) {\n \t\tldr := ctxt.loader\n \n \t\t// must match domacholink below\n-\t\ts1 := ldr.SymSize(ldr.Lookup(\".machosymtab\", 0))\n-\t\ts2 := ldr.SymSize(ctxt.ArchSyms.LinkEditPLT)\n-\t\ts3 := ldr.SymSize(ct...
2020-11-12T00:43:26
nodejs/node
86d386a07d3bb7d13c0b174f37cf2bc1d00dee51
68726900437cf07861cc2439d81a3842ce7a3b3e
doc: fix typo in dns.md PR-URL: https://github.com/nodejs/node/pull/17502 Reviewed-By: Colin Ihrig <cjihrig@gmail.com> Reviewed-By: Richard Lau <riclau@uk.ibm.com> Reviewed-By: Anatoli Papirovski <apapirovski@mac.com> Reviewed-By: Jon Moss <me@jonathanmoss.me> Reviewed-By: Minwoo Jung <minwoo@nodesource.com> Reviewed-...
[ { "path": "doc/api/dns.md", "patch": "@@ -514,7 +514,7 @@ will be present on the object:\n | `\"SRV\"` | Refer to [`dns.resolveSrv()`][] |\n | `\"TXT\"` | This type of record contains an array property called `entries` which refers to [`dns.resolveTxt()`][], eg. `{ entries: ['...'], type: 'TXT' }` |\n \n-He...
2017-12-06T18:13:20
ggml-org/llama.cpp
b8635075ffe27b135c49afb9a8b5c434bd42c502
9c699074c97191754c8a966298f84c79f90fce38
common : add gemma 4 specialized parser (#21418) * common : add gemma4 dedicated parser * cont : add '<|tool_response>' as eog * cont : emit JSON from Gemma4 tool call AST * cont : more fixes * cont : refactor convert function * cont : refine rules and mapping * cont : add more tests * cont : clean up * cont :...
[ { "path": "common/chat-auto-parser-generator.cpp", "patch": "@@ -8,109 +8,11 @@\n #include \"nlohmann/json.hpp\"\n #include \"peg-parser.h\"\n \n-#include <algorithm>\n #include <stdexcept>\n #include <string>\n \n using json = nlohmann::ordered_json;\n \n-namespace {\n-\n-// Gemma4-specific PEG builder ext...
2026-04-04T18:39:00
huggingface/transformers
01be5b48790f113b7d71943b580c842e3e097988
c85510f958e6955d88ea1bafb4f320074bfbd0c1
adds: extra_repr() to MambaRMSNorm to include hidden size / size of weights in the layer (#32171) * adds: extra_repr() to MambaRMSNorm to include the hidden size of the layer * style fix with ruff:
[ { "path": "src/transformers/models/mamba/modeling_mamba.py", "patch": "@@ -327,6 +327,9 @@ def forward(self, hidden_states):\n hidden_states = hidden_states * torch.rsqrt(variance + self.variance_epsilon)\n return self.weight * hidden_states.to(input_dtype)\n \n+ def extra_repr(self):\n+ ...
2024-07-24T07:09:59
golang/go
f2e58c6d4239f27db284dfe442fa62bb3c0c5b23
4c174a7ba66724f8f9a1915c8f4868a8b3aaf219
syscall: improve TestSetuidEtc() /proc/ parsing against races TestSetuidEtc() was failing sporadically on linux-ppc64. From the three https://build.golang.org/ logs, it looked like the logged errors could be associated with threads dying, but proc reads were, in some way, racing with their demise. Exploring ways to i...
[ { "path": "misc/cgo/test/issue1435.go", "patch": "@@ -62,28 +62,60 @@ import \"C\"\n // compareStatus is used to confirm the contents of the thread\n // specific status files match expectations.\n func compareStatus(filter, expect string) error {\n-\texpected := filter + \"\\t\" + expect\n+\texpected := fil...
2020-11-10T03:28:04
nodejs/node
d269e07fa49952d7b414cee8636626bf2ac0affa
af446a98b9402180116f020cf76dbb5b23270234
deps: add -no_rand_screen to openssl s_client In openssl s_client on Windows, RAND_screen() is invoked to initialize random state but it takes several seconds in each connection. This added -no_rand_screen to openssl s_client on Windows to skip RAND_screen() and gets a better performance in the unit test of test-tls-s...
[ { "path": "deps/openssl/openssl/apps/app_rand.c", "patch": "@@ -124,7 +124,16 @@ int app_RAND_load_file(const char *file, BIO *bio_e, int dont_warn)\n char buffer[200];\n \n #ifdef OPENSSL_SYS_WINDOWS\n- RAND_screen();\n+ /*\n+ * allocate 2 to dont_warn not to use RAND_screen() via\n+ * -n...
2015-05-27T01:33:38
ggml-org/llama.cpp
9c699074c97191754c8a966298f84c79f90fce38
d01f6274c01e111be2ccc39443f79884796e48fb
server: Fix undefined timing measurement errors in server context (#21201) Co-authored-by: Dan Hoffman <dhoffman@cyket.net>
[ { "path": "tools/server/server-context.cpp", "patch": "@@ -155,8 +155,8 @@ struct server_slot {\n int64_t t_start_process_prompt;\n int64_t t_start_generation;\n \n- double t_prompt_processing; // ms\n- double t_token_generation; // ms\n+ double t_prompt_processing = 0.0; // ms\n+ doubl...
2026-04-04T14:11:19
huggingface/transformers
c85510f958e6955d88ea1bafb4f320074bfbd0c1
bc2adb0112b6677b0dfb4105c74570a0f92183eb
[docs] change temperature to a positive value (#32077) fix
[ { "path": "docs/source/en/conversations.md", "patch": "@@ -195,7 +195,7 @@ inputs = {key: tensor.to(model.device) for key, tensor in inputs.items()}\n print(\"Tokenized inputs:\\n\", inputs)\n \n # 4: Generate text from the model\n-outputs = model.generate(**inputs, max_new_tokens=512, temperature=0.)\n+out...
2024-07-23T16:47:51
ollama/ollama
7bdcd1da9417ecc1a1069f2d73f0b88a3a43857d
0f3cf1d42e47f5293c14422032c324a894e3dd93
Revert "Merge pull request #4938 from ollama/mxyng/fix-byte-order" This reverts commit f5f245cc154580fa7b4052c001d2a7e3d771cfb8, reversing changes made to 94d37fdcae30ddeb6c9f65c8707004f5ec9eaf33. this change broke gguf v2 which is incorrectly detected as big endian
[ { "path": "llm/ggml.go", "patch": "@@ -231,7 +231,8 @@ const (\n \t// Magic constant for `ggla` files (LoRA adapter).\n \tFILE_MAGIC_GGLA = 0x67676C61\n \t// Magic constant for `gguf` files (versioned, gguf)\n-\tFILE_MAGIC_GGUF = 0x46554747\n+\tFILE_MAGIC_GGUF_LE = 0x46554747\n+\tFILE_MAGIC_GGUF_BE = 0x4747...
2024-06-11T22:55:44