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
nodejs/node
5856c836eaa2738269470e16b7ac7c1a92eac6d0
fa939f0cf59fdafae6748b390bcfc733f188b425
src: fix vm module for strict mode This patch fixes the problem with variables that are declared only on the sandbox but not on the global proxy. PR-URL: https://github.com/nodejs/node/pull/16487 Fixes: https://github.com/nodejs/node/issues/12300 Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl> Reviewed-By: Michaël Za...
[ { "path": "src/node_contextify.cc", "patch": "@@ -346,14 +346,21 @@ class ContextifyContext {\n return;\n \n auto attributes = PropertyAttribute::None;\n- bool is_declared = ctx->global_proxy()\n+ bool is_declared_on_global_proxy = ctx->global_proxy()\n ->GetRealNamedPropertyAttribut...
2017-10-25T11:49:58
rust-lang/rust
6c1e5b4edba347920c44e94b80d07d298c9104a7
768749f41dcfb11a77a160de704cecc83be28a0f
Fix transparent diagnostics
[ { "path": "src/tools/rust-analyzer/editors/code/src/diagnostics.ts", "patch": "@@ -207,11 +207,6 @@ export class AnsiDecorationProvider implements vscode.Disposable {\n }\n }\n \n- const themeColor = AnsiDecorationProvider._anserToThemeColor[color];\n- if (themeColor) {\n- ...
2025-03-02T14:05:13
ggml-org/llama.cpp
423bee462b46323433e5f1e322e3076cedbeec83
8abcc70a747eff568198ae64aa1a60b7625a3c36
ci : fix sanitize workflow to enable ggml sanitizers too (#19323)
[ { "path": ".github/workflows/build.yml", "patch": "@@ -293,6 +293,7 @@ jobs:\n cmake -B build \\\n -DLLAMA_FATAL_WARNINGS=ON \\\n -DLLAMA_SANITIZE_${{ matrix.sanitizer }}=ON \\\n+ -DGGML_SANITIZE_${{ matrix.sanitizer }}=ON \\\n -DCMAKE_BUILD_TYPE=${{ ...
2026-02-04T13:12:03
golang/go
1eeaff75f9e02c65d29d9910c1884c6c0ecc1430
ae329abec0f78743ab2fbf30ef5b488376fe3c85
crypto/x509: enforce SAN IA5String encoding restrictions Extends the IA5String encoding restrictions that are currently applied to name constraints to dNSName, rfc822Name, and uniformResourceIdentifier elements of the SAN. The utility function isIA5String is updated to use unicode.MaxASCII rather than utf8.RuneSelf as...
[ { "path": "doc/go1.16.html", "patch": "@@ -174,6 +174,16 @@ <h3 id=\"ctypto/tls\"><a href=\"/pkg/crypto/tls\">crypto/tls</a></h3>\n by the <code>Error</code> method with <code>\"tls: use of closed connection\"</code>.\n </p>\n \n+<h3 id=\"crypto/x509\"><a href=\"/pkg/crypto/x509\">crypto/x509</a></h3>\n+\...
2020-05-23T17:15:46
huggingface/transformers
bd9f4d79517a3ad2f9da999d090dc3bbfc506dc4
b8aee2e918d7ba2d5e9e80162ae26b4806873307
Add Video Llava (#29733) * add model draft * update docstring * add tests * support image and video as input * update for better handling of mixed input and clean-up a bit * bug when mixed inputs & add tests * Update README.md Co-authored-by: NielsRogge <48327001+NielsRogge@users.noreply.github.c...
[ { "path": "README_fr.md", "patch": "@@ -288,7 +288,6 @@ Suivez les pages d'installation de Flax, PyTorch ou TensorFlow pour voir comment\n \n Nombre actuel de points de contrôle : ![](https://img.shields.io/endpoint?url=https://huggingface.co/api/shields/models&color=brightgreen)\n \n-\n 🤗 Transformers fou...
2024-05-15T11:42:29
ollama/ollama
0b3118e0afe1a4658264081979b04aab9fda82d6
05face44efdb06d978394aadfce98daae7dca8a8
fix: relay request opts to loaded llm prediction (#1761)
[ { "path": "llm/ext_server_common.go", "patch": "@@ -153,7 +153,7 @@ func newExtServer(server extServer, model string, adapters, projectors []string,\n \treturn server, nil\n }\n \n-func predict(llm extServer, opts api.Options, ctx context.Context, predict PredictOpts, fn func(PredictResult)) error {\n+func ...
2024-01-03T17:01:42
nodejs/node
2be9198097a19169026fb73e263165e6cbee7243
748f5d848f229f646850401e644818bc5599ec9b
test: change tmp directories prefix `.tmp` prefix allows easier exclusion PR-URL: https://github.com/nodejs/node/pull/16372 Refs: https://github.com/nodejs/node/pull/16010 Refs: https://github.com/nodejs/node/issues/16278 Reviewed-By: Rich Trott <rtrott@gmail.com> Reviewed-By: Joyee Cheung <joyeec9h3@gmail.com> Revie...
[ { "path": ".eslintignore", "patch": "@@ -2,7 +2,6 @@ lib/internal/v8_prof_polyfill.js\n lib/punycode.js\n test/addons/??_*\n test/fixtures\n-test/tmp*\n tools/eslint\n tools/icu\n node_modules", "additions": 0, "deletions": 1, "language": "Unknown" }, { "path": "test/common/index.js", ...
2017-10-21T22:07:19
ggml-org/llama.cpp
d838c22bb33d6b486a73c0ff285346deb575be3f
25f40ca65f1aa596f8b1702bbac4bc48a45b87d7
spec : fix the check-rate logic of ngram-simple (#19261) * spec : fix the check-rate logic of ngram-simple * cont : refactor + fix checks
[ { "path": "common/ngram-map.cpp", "patch": "@@ -47,21 +47,15 @@ static std::string common_tokens_to_str(const llama_tokens & inp, size_t start,\n * @return Vector of draft tokens, empty if no matching pattern is found\n */\n llama_tokens common_ngram_simple_draft(\n- common_ngram_simple_state & sta...
2026-02-04T08:39:53
golang/go
27280d8c14331c1c46cd90206be9f3c924f6b4c4
54a112d7197ec320527614e7502a3243eab93b6e
crypto/x509: return errors instead of panicking Eliminate a panic in x509.CreateCertificate when passing templates with unknown ExtKeyUsage; return an error instead. Fixes #41169 Change-Id: Ia229d3b0d4a1bdeef05928439d97dab228687b3c Reviewed-on: https://go-review.googlesource.com/c/go/+/252557 Reviewed-by: Roland Sho...
[ { "path": "src/crypto/x509/x509.go", "patch": "@@ -1689,7 +1689,8 @@ func buildExtensions(template *Certificate, subjectIsEmpty bool, authorityKeyId\n \t\t\tif oid, ok := oidFromExtKeyUsage(u); ok {\n \t\t\t\toids = append(oids, oid)\n \t\t\t} else {\n-\t\t\t\tpanic(\"internal error\")\n+\t\t\t\terr = error...
2020-09-02T10:44:36
huggingface/transformers
b8aee2e918d7ba2d5e9e80162ae26b4806873307
be3aa43e5fd207533136a380a43322a83af53846
Remove unused module DETR based models (#30823) * removing heads for classification from DETR models. * quality fix
[ { "path": "src/transformers/models/conditional_detr/modeling_conditional_detr.py", "patch": "@@ -1091,25 +1091,6 @@ def forward(\n return outputs\n \n \n-# Copied from transformers.models.detr.modeling_detr.DetrClassificationHead with Detr->ConditionalDetr\n-class ConditionalDetrClassificationHead(n...
2024-05-15T10:19:43
rust-lang/rust
ef028d79d6d4d54576c88dbd1f9de7f4824c0289
ad27045c31a9f37ad7d44ca2a403de52d1a896d3
do not build additional stage on compiler paths When calling `x build compiler (or rustc) --stage N` bootstrap builds stage N+1 compiler, which is clearly not what we requested. This doesn't happen when running `x build --stage N` without explicitly targeting the compiler. The changes applied fix this issue. Signed-...
[ { "path": "src/bootstrap/src/core/build_steps/compile.rs", "patch": "@@ -993,7 +993,9 @@ impl Step for Rustc {\n fn make_run(run: RunConfig<'_>) {\n let crates = run.cargo_crates_in_set();\n run.builder.ensure(Rustc {\n- compiler: run.builder.compiler(run.builder.top_stage, ru...
2025-03-02T08:27:40
nodejs/node
b61a08ca2d1e6120256e939381361cc0ce5a28a9
0a03e350fb6fac67ac85809af947c8dce1a8e569
http2: fix mapToHeaders() with single string value This is for issue 16452. When 'set-cookie' header is set with an array that has only one string value, it's split into its individual characters. Fix by resetting `isArray` to false when the value is converted from an array to a string. Fixes: https://github.com/nod...
[ { "path": "lib/internal/http2/util.js", "patch": "@@ -403,13 +403,14 @@ function mapToHeaders(map,\n if (typeof key === 'symbol' || value === undefined || !key)\n continue;\n key = String(key).toLowerCase();\n- const isArray = Array.isArray(value);\n+ let isArray = Array.isArray(value);\...
2017-10-24T20:34:25
ggml-org/llama.cpp
015deb90485b9ebfff492e10fe7080a0439202b6
2ceda3f6622661af839c19767705f33fb9f6cdd2
ggml-virtgpu: make the code thread safe (#19204) * ggml-virtgpu: regenerate_remoting.py: add the ability to deprecate a function * ggml-virtgpu: deprecate buffer_type is_host remoting not necessary * ggml-virtgpu: stop using static vars as cache The static init isn't thread safe. * ggml-virtgpu: protect the use o...
[ { "path": "ggml/include/ggml-virtgpu.h", "patch": "@@ -7,8 +7,6 @@\n extern \"C\" {\n #endif\n \n-#define GGML_REMOTING_FRONTEND_NAME \"RemotingFrontend\"\n-\n GGML_BACKEND_API ggml_backend_reg_t ggml_backend_virtgpu_reg();\n \n #ifdef __cplusplus", "additions": 0, "deletions": 2, "language": "...
2026-02-04T02:46:18
huggingface/transformers
be3aa43e5fd207533136a380a43322a83af53846
37543bad3c1589bfe469e76f896b7fd5e5b1d0e6
Support mixed-language batches in `WhisperGenerationMixin` (#29688) * Add support for mixing languages in a single batch * Update docstring * Enable different detected languages in batch * Do not require input_features * Test list of languages * Fix comment * Make init_tokens length-1 if possible, br...
[ { "path": "src/transformers/models/whisper/generation_whisper.py", "patch": "@@ -262,7 +262,7 @@ def generate(\n synced_gpus: bool = False,\n return_timestamps: Optional[bool] = None,\n task: Optional[str] = None,\n- language: Optional[str] = None,\n+ language: Optional...
2024-05-15T07:53:17
golang/go
8b0d00b1645c47076f5b20dc692b2ca6d9bac19b
3caaaddffd605c0ec1787d68295b732fff73026b
cmd/go: error if -modfile used with 'go install pkg@version' 'go install pkg@version' runs without a main module or a module root directory. The -modfile flag cannot be used to set the module root directory or to substitute a different go.mod file. This error won't be reported if -modfile is set in GOFLAGS. Unsupport...
[ { "path": "src/cmd/go/internal/modload/init.go", "patch": "@@ -163,8 +163,9 @@ func Init() {\n \t\t// Running 'go mod init': go.mod will be created in current directory.\n \t\tmodRoot = base.Cwd\n \t} else if RootMode == NoRoot {\n-\t\t// TODO(jayconrod): report an error if -mod -modfile is explicitly set o...
2020-09-29T20:40:57
ollama/ollama
a2ad952440feaf997cfdc9f0d81d0f545bec4974
5fea4410be0420fabfb6c75390ad8081d49fca42
Fix windows system memory lookup This refines the gpu package error handling and fixes a bug with the system memory lookup on windows.
[ { "path": "gpu/gpu.go", "patch": "@@ -66,7 +66,7 @@ func GetGPUInfo() GpuInfo {\n \t}\n \n \tvar memInfo C.mem_info_t\n-\tresp := GpuInfo{\"\", 0, 0}\n+\tresp := GpuInfo{}\n \tif gpuHandles.cuda != nil {\n \t\tC.cuda_check_vram(*gpuHandles.cuda, &memInfo)\n \t\tif memInfo.err != nil {\n@@ -103,6 +103,19 @@ ...
2023-12-22T23:43:31
rust-lang/rust
5a2da96a44abad6be752c7ee1cac173aa1ca2f7c
a0ed304c2101b4efdaf9982189e8d6ca73690712
dec2flt: Update documentation of existing methods Fix or elaborate existing float parsing documentation. This includes introducing a convention that should make naming more consistent.
[ { "path": "library/core/src/num/dec2flt/common.rs", "patch": "@@ -8,12 +8,12 @@ pub(crate) trait ByteSlice {\n /// Writes a 64-bit integer as 8 bytes in little-endian order.\n fn write_u64(&mut self, value: u64);\n \n- /// Calculate the offset of a slice from another.\n+ /// Calculate the diff...
2024-12-09T09:25:22
ggml-org/llama.cpp
8bece2eb20f0134632ae229849fbde6559882d36
a6fd8ca1fee621addff1695165414c4822fb08bf
CUDA: use mmvq for mul-mat-id for small batch sizes (#18958) * CUDA: use mmvq for mul-mat-id for small batch sizes * add mmvq too * Fix perf issue on ampere. Use mmvf mm-id only for non-nvidia GPUs * templatize multi_token_path
[ { "path": "ggml/src/ggml-cuda/ggml-cuda.cu", "patch": "@@ -2279,13 +2279,19 @@ static void ggml_cuda_mul_mat_id(ggml_backend_cuda_context & ctx, ggml_tensor *\n const int cc = ggml_cuda_info().devices[ggml_cuda_get_device()].cc;\n \n if (src1->type == GGML_TYPE_F32 && dst->type == GGML_TYPE_F32) {\n...
2026-02-03T15:31:23
huggingface/transformers
37543bad3c1589bfe469e76f896b7fd5e5b1d0e6
99e16120ab46216bb7b989f69737ce8cd4ba7677
Add missing dependencies in image classification example (#30820) fix: missing dependencies
[ { "path": "docs/source/en/tasks/image_classification.md", "patch": "@@ -38,7 +38,7 @@ To see all architectures and checkpoints compatible with this task, we recommend\n Before you begin, make sure you have all the necessary libraries installed:\n \n ```bash\n-pip install transformers datasets evaluate accel...
2024-05-15T06:38:30
nodejs/node
0a03e350fb6fac67ac85809af947c8dce1a8e569
76b8803630f9a324810f38d00610792033e862c7
crypto: migrate crypto.randomBytes to internal/errors PR-URL: https://github.com/nodejs/node/pull/16454 Reviewed-By: Michaël Zasso <targos@protonmail.com> Reviewed-By: Anatoli Papirovski <apapirovski@mac.com> Reviewed-By: Joyee Cheung <joyeec9h3@gmail.com> Reviewed-By: Luigi Pinca <luigipinca@gmail.com>
[ { "path": "lib/internal/crypto/random.js", "patch": "@@ -3,7 +3,7 @@\n const errors = require('internal/errors');\n const { isArrayBufferView } = require('internal/util/types');\n const {\n- randomBytes,\n+ randomBytes: _randomBytes,\n randomFill: _randomFill\n } = process.binding('crypto');\n \n@@ -24,...
2017-10-24T19:26:03
golang/go
3caaaddffd605c0ec1787d68295b732fff73026b
0e85fd7561de869add933801c531bf25dee9561c
runtime: don't crash if vsyscall and vdso are disabled on x86_64 If vdso is disabled, the goruntime calls gettimeofday from vsyscall, but if vsyscall is disabled too, all golang binaries crash: SIGSEGV {si_signo=SIGSEGV, si_code=SEGV_MAPERR, si_addr=0xffffffffff600000} --- killed by SIGSEGV (core dumped) ++ vsyscall...
[ { "path": "src/runtime/sys_linux_amd64.s", "patch": "@@ -40,6 +40,7 @@\n #define SYS_futex\t\t202\n #define SYS_sched_getaffinity\t204\n #define SYS_epoll_create\t213\n+#define SYS_clock_gettime\t228\n #define SYS_exit_group\t\t231\n #define SYS_epoll_ctl\t\t233\n #define SYS_tgkill\t\t234\n@@ -241,15 +242,...
2020-09-29T04:45:55
ollama/ollama
b846eb64d0df18e472ea52383d8c0c2078815da2
3c5dd9ed1d66e3f96ed75bf9162b29e7b923b018
Fix `template` api doc description (#1661)
[ { "path": "docs/api.md", "patch": "@@ -46,7 +46,7 @@ Advanced parameters (optional):\n - `format`: the format to return a response in. Currently the only accepted value is `json`\n - `options`: additional model parameters listed in the documentation for the [Modelfile](./modelfile.md#valid-parameters-and-va...
2024-01-03T16:00:59
ggml-org/llama.cpp
1f1e57f2bf76df88a668dce924de2bffa89c5c02
e9a859db3cbad0d3dd10deb0527ddfcecce5e78d
CUDA: Fix loop unrolling for BW in mul_mat_q_stream_k_fixup (#19053) By providing stride_* variables as size_t (i.e., 64-bit) the compiler can correctly unroll the [two for-loops](https://github.com/ggml-org/llama.cpp/blob/557515be1e93ed8939dd8a7c7d08765fdbe8be31/ggml/src/ggml-cuda/mmq.cuh#L3789-L3816) on BW. This giv...
[ { "path": "ggml/src/ggml-cuda/mmq.cuh", "patch": "@@ -3697,13 +3697,20 @@ static __global__ void mul_mat_q(\n tile_x_max_i, tile_y_max_j, kb0_start, kb0_stop);\n }\n \n-\n template <ggml_type type, int mmq_x, bool need_check>\n-static __global__ void mul_mat_q_stream_k_fixup(\n- const int32_...
2026-02-03T10:33:14
huggingface/transformers
99e16120ab46216bb7b989f69737ce8cd4ba7677
1360801a69c0b169e3efdbb0cd05d9a0e72bfb70
Add support for custom checkpoints in MusicGen (#30011) * feat: support custom checkpoint * update: revert changes and add TODO * update: docs and exception handling * fix: ah, extra space
[ { "path": "src/transformers/models/musicgen/convert_musicgen_transformers.py", "patch": "@@ -88,24 +88,24 @@ def rename_state_dict(state_dict: OrderedDict, hidden_size: int) -> Tuple[Dict,\n \n \n def decoder_config_from_checkpoint(checkpoint: str) -> MusicgenDecoderConfig:\n- if checkpoint == \"small\" ...
2024-05-15T06:30:33
nodejs/node
76b8803630f9a324810f38d00610792033e862c7
fb477f3fa5c944e24b3b8275e4986ea3319b63a0
tools: add eslint rule for documented errors PR-URL: https://github.com/nodejs/node/pull/16450 Reviewed-By: Anatoli Papirovski <apapirovski@mac.com> Reviewed-By: Refael Ackermann <refack@gmail.com> Reviewed-By: Colin Ihrig <cjihrig@gmail.com> Reviewed-By: Teddy Katz <teddy.katz@gmail.com>
[ { "path": "lib/internal/errors.js", "patch": "@@ -1,3 +1,4 @@\n+/* eslint documented-errors: \"error\" */\n /* eslint alphabetize-errors: \"error\" */\n \n 'use strict';", "additions": 1, "deletions": 0, "language": "JavaScript" }, { "path": "test/parallel/test-eslint-documented-errors.j...
2017-10-24T17:57:21
golang/go
0e85fd7561de869add933801c531bf25dee9561c
af9c5e5dbc3a5abc49aa3ac45da1b533b0d238a6
cmd/compile: report type loop for invalid recursive types Similar to how we report initialization loops in initorder.go and type alias loops in typecheck.go, this CL updates align.go to warn about invalid recursive types. The code is based on the loop code from initorder.go, with minimal changes to adapt from detectin...
[ { "path": "src/cmd/compile/internal/gc/align.go", "patch": "@@ -5,7 +5,9 @@\n package gc\n \n import (\n+\t\"bytes\"\n \t\"cmd/compile/internal/types\"\n+\t\"fmt\"\n \t\"sort\"\n )\n \n@@ -173,6 +175,91 @@ func widstruct(errtype *types.Type, t *types.Type, o int64, flag int) int64 {\n \treturn o\n }\n \n+//...
2020-09-29T09:11:10
ollama/ollama
e5202eb6879a8c29b5b4646ac0d4d93e1b0d1c54
96fb441abd297748cfddb173289af7d54dbeee4d
Quiet down llama.cpp logging by default By default builds will now produce non-debug and non-verbose binaries. To enable verbose logs in llama.cpp and debug symbols in the native code, set `CGO_CFLAGS=-g`
[ { "path": "docs/development.md", "patch": "@@ -14,7 +14,13 @@ Install required tools:\n brew install go cmake gcc\n ```\n \n-Get the required libraries:\n+Optionally enable debugging and more verbose logging:\n+\n+```bash\n+export CGO_CFLAGS=\"-g\"\n+```\n+\n+Get the required libraries and build the native ...
2023-12-22T16:47:18
huggingface/transformers
1360801a69c0b169e3efdbb0cd05d9a0e72bfb70
c96aca3a8d66d64f868a3e3967be624d79213bef
Add PaliGemma (#30814) * add new model like * add state dict slicing + new model config * update palma config and weights, passes vision activations * fix * update * reorder loading/unpacking * clean up * add debug statements * change device * fix * debugging * fix noncausal mask * ...
[ { "path": "docs/source/en/_toctree.yml", "patch": "@@ -784,6 +784,8 @@\n title: OWL-ViT\n - local: model_doc/owlv2\n title: OWLv2\n+ - local: model_doc/paligemma\n+ title: PaliGemma\n - local: model_doc/perceiver\n title: Perceiver\n - local: model_doc/p...
2024-05-14T20:07:15
nodejs/node
fb477f3fa5c944e24b3b8275e4986ea3319b63a0
eeada6ca63d57354aef3e5907097c56a5b9fb347
doc: add missing error codes PR-URL: https://github.com/nodejs/node/pull/16450 Reviewed-By: Anatoli Papirovski <apapirovski@mac.com> Reviewed-By: Refael Ackermann <refack@gmail.com> Reviewed-By: Colin Ihrig <cjihrig@gmail.com> Reviewed-By: Teddy Katz <teddy.katz@gmail.com>
[ { "path": "doc/api/errors.md", "patch": "@@ -643,11 +643,36 @@ Used when an invalid value for the `format` argument has been passed to the\n Used when an invalid crypto engine identifier is passed to\n [`require('crypto').setEngine()`][].\n \n+<a id=\"ERR_CRYPTO_HASH_DIGEST_NO_UTF16\"></a>\n+### ERR_CRYPTO_...
2017-10-24T17:17:53
ggml-org/llama.cpp
91ea44e89b30474831d5dc0ad57719a5819506db
0dfcd3b60755bf9cb3c9ec726a584b8a4f20239b
opencl: refactor some ops, concat, repeat, tanh and scale (#19226) * opencl: refactor concat * opencl: refactor repeat * opencl: refactor tanh * opencl: enable fp16 for tanh * opencl: refactor scale * opencl: fix unused variables
[ { "path": "ggml/src/ggml-opencl/ggml-opencl.cpp", "patch": "@@ -453,7 +453,6 @@ struct ggml_backend_opencl_context {\n cl_program program_rms_norm;\n cl_program program_group_norm;\n cl_program program_rope;\n- cl_program program_scale;\n cl_program program_silu;\n cl_program program_...
2026-02-02T23:54:43
golang/go
af9c5e5dbc3a5abc49aa3ac45da1b533b0d238a6
567ef8bd8e76bdbc00df6b1903976b89b34a84d8
crypto/x509: prioritize potential parents in chain building When building a x509 chain the algorithm currently looks for parents that have a subject key identifier (SKID) that matches the child authority key identifier (AKID), if it is present, and returns all matches. If the child doesn't have an AKID, or there are n...
[ { "path": "src/crypto/x509/cert_pool.go", "patch": "@@ -5,36 +5,29 @@\n package x509\n \n import (\n+\t\"bytes\"\n \t\"encoding/pem\"\n \t\"errors\"\n \t\"runtime\"\n )\n \n // CertPool is a set of certificates.\n type CertPool struct {\n-\tbySubjectKeyId map[string][]int\n-\tbyName map[string][]int...
2020-05-08T22:57:25
ollama/ollama
495c06e4a67ecc1171faa22115443103a24af1da
fa24e73b8253a554ec840395a5d1dfdb91d3598b
Fix doc glitch
[ { "path": "docs/modelfile.md", "patch": "@@ -188,7 +188,7 @@ SYSTEM \"\"\"<system message>\"\"\"\n \n ### ADAPTER\n \n-The `ADAPTER` instruction specifies the LoRA adapter to apply to the base model. The value of this instruction should be an absolute path or a path relative to the Modelfile and the file mu...
2023-12-22T00:57:58
rust-lang/rust
c51b229140c885cac757a405a328a07e90d5bca9
8c392966a013fd8a09e6b78b3c8d6e442bc278e1
Disable `f16` on Aarch64 without `neon` LLVM has crashes at some `half` operations when built with assertions enabled if fp-armv8 is not available [1]. Things seem to usually work, but we are reaching LLVM undefined behavior so this needs to be disabled. [1]: https://github.com/llvm/llvm-project/issues/129394
[ { "path": "library/std/build.rs", "patch": "@@ -12,6 +12,11 @@ fn main() {\n .expect(\"CARGO_CFG_TARGET_POINTER_WIDTH was not set\")\n .parse()\n .unwrap();\n+ let target_features: Vec<_> = env::var(\"CARGO_CFG_TARGET_FEATURE\")\n+ .unwrap_or_default()\n+ .split(\",\...
2025-03-01T23:05:19
huggingface/transformers
ccdabc5642bf84849af93f591e207dc625c8e1e1
d84f34ad7744bd713428550806c71959e38b1097
Add JetMoE model (#30005) * init jetmoe code * update archive maps * remove flax import * fix import error * update README * ruff fix * update readme * fix * update config * fix issue * merge files * fix model bug * fix test * auto fix * model size * add comments * fix f...
[ { "path": "docs/source/en/_toctree.yml", "patch": "@@ -386,6 +386,8 @@\n title: I-BERT\n - local: model_doc/jamba\n title: Jamba\n+ - local: model_doc/jetmoe\n+ title: JetMoe\n - local: model_doc/jukebox\n title: Jukebox\n - local: model_doc/led", "a...
2024-05-14T14:32:01
ggml-org/llama.cpp
a3fa03582240a4279ba019a3db2bb87311d5d485
15818ac44c551d6d668c69bcfbf9a937c98f9e5a
server: print actual model name in 'model not found" error (#19117) Experimenting with AI, my environment gets messy fast and it's not always easy to know what model my software is trying to load. This helps with troubleshooting. before: Error: { code = 400, message = "model not found", type = "invalid_request...
[ { "path": "tools/server/server-models.cpp", "patch": "@@ -767,7 +767,7 @@ static bool router_validate_model(const std::string & name, server_models & mode\n }\n auto meta = models.get_meta(name);\n if (!meta.has_value()) {\n- res_err(res, format_error_response(\"model not found\", ERROR_T...
2026-02-02T15:55:27
nodejs/node
eeada6ca63d57354aef3e5907097c56a5b9fb347
664512678d8d02e2ef413136cf262dcf7debb42d
crypto: migrate timingSafeEqual to internal/errors PR-URL: https://github.com/nodejs/node/pull/16448 Reviewed-By: Michaël Zasso <targos@protonmail.com> Reviewed-By: Joyee Cheung <joyeec9h3@gmail.com>
[ { "path": "doc/api/errors.md", "patch": "@@ -648,6 +648,12 @@ Used when an invalid crypto engine identifier is passed to\n \n Used when an invalid [crypto digest algorithm][] is specified.\n \n+<a id=\"ERR_CRYPTO_TIMING_SAFE_EQUAL_LENGTH\"></a>\n+### ERR_CRYPTO_TIMING_SAFE_EQUAL_LENGTH\n+\n+Used when callin...
2017-10-24T17:04:25
golang/go
567ef8bd8e76bdbc00df6b1903976b89b34a84d8
66770f4b1de37d9c5c962edb2980a70102e09ec3
runtime: correct arguments to pipe2 syscall in pipe on netbsd/arm64 Follow the implementation of pipe2 to pass arguments and handle return values. With this fix the runtime tests pass on netbsd/arm64 (tested using gomote on the netbsd-arm64-bsiegert builder). Update #30824 Change-Id: I346db68fc9dde8dc7f11351af05c1a...
[ { "path": "src/runtime/sys_netbsd_arm64.s", "patch": "@@ -152,19 +152,13 @@ ok:\n \n // func pipe() (r, w int32, errno int32)\n TEXT runtime·pipe(SB),NOSPLIT|NOFRAME,$0-12\n-\tMOVW\t$0, R0\n+\tADD\t$8, RSP, R0\n+\tMOVW\t$0, R1\n \tSVC\t$SYS_pipe2\n \tBCC\tpipeok\n-\tMOVW\t$-1,R1\n-\tMOVW\tR1, r+0(FP)\n-\tMO...
2020-09-29T13:13:17
ollama/ollama
fa24e73b8253a554ec840395a5d1dfdb91d3598b
325d74985b9f31917ead1585ea22389a39b280b5
Remove CPU build, fixup linux build script
[ { "path": "Dockerfile.build", "patch": "@@ -4,6 +4,7 @@ ARG CUDA_VERSION=11.3.1-1\n ARG CMAKE_VERSION=3.22.1\n # ROCm only supports amd64\n ARG ROCM_VERSION=6.0\n+ARG CLBLAST_VER=1.6.1\n \n # Note: https://rocm.docs.amd.com/en/latest/release/user_kernel_space_compat_matrix.html\n RUN apt-get update && \\\n@...
2023-12-22T00:54:54
huggingface/transformers
d84f34ad7744bd713428550806c71959e38b1097
9ef38840468ad5d71c3f1380da627d9ecde439d1
[T5] Adding `model_parallel = False` to `T5ForTokenClassification` and `MT5ForTokenClassification` (#30763) * Adding model_parallel = False * Revert "Adding model_parallel = False" This reverts commit ba1d99976acb598824ce3347dbe7d848daa21e79. * Trainer: circumvent error for model in which is_parallelizable i...
[ { "path": "src/transformers/trainer.py", "patch": "@@ -436,7 +436,7 @@ def __init__(\n \"https://huggingface.co/docs/transformers/model_doc/auto\"\n )\n \n- if hasattr(model, \"is_parallelizable\") and model.is_parallelizable and model.model_parallel:\n+ if getattr(...
2024-05-14T13:39:25
ggml-org/llama.cpp
6156ae51114337ffb6bb46cb65f99227b255089f
59377a6c870be95e4c71715933e4e9ada71b8356
model-conversion : add debug option to conversion script (#19265) This commit adds a debug option to the model conversion script to enable using the Python debugger (pdb) during model conversion. The motivation for this is that I've found myself adding this a few times now and it would be quicker to have this flag as...
[ { "path": "examples/model-conversion/Makefile", "patch": "@@ -33,11 +33,14 @@ DEVICE ?= auto\n causal-convert-model-bf16: OUTTYPE=bf16\n causal-convert-model-bf16: causal-convert-model\n \n+causal-convert-model-debug: DEBUG=--debug\n+causal-convert-model-debug: causal-convert-model\n+\n causal-convert-model...
2026-02-02T10:29:57
nodejs/node
d653f03c0e9fa097abb9959931a78d02a3eb8814
36732084db9d0ff59b6ce31e839450cd91a156be
test: allow tests to pass without internet Currently when running the test without an internet connection there are two JavaScript test failures and one cctest. The cctest only fails on Mac as far as I know. (I've only tested using Mac and Linux thus far). This commit moves the two JavaScript tests to test/internet. ...
[ { "path": "test/cctest/test_inspector_socket_server.cc", "patch": "@@ -589,7 +589,7 @@ TEST_F(InspectorSocketServerTest, TerminatingSessionReportsDone) {\n \n TEST_F(InspectorSocketServerTest, FailsToBindToNodejsHost) {\n TestInspectorServerDelegate delegate;\n- ServerHolder server(&delegate, &loop, \"no...
2017-10-17T05:27:07
golang/go
0ab72ed020d0c320b5007987abdf40677db34cfc
0163bdae685c1b060f8108ac5af13ea6374555b1
cmd/link, runtime: use a sentinel value for unreachable method In the method table, the method's code pointer is stored as an offset from the start of the text section. Currently, for an unreachable method, the offset is left as 0, which resolves to the start of the text section at run time. It is possible that there ...
[ { "path": "src/cmd/link/internal/ld/data.go", "patch": "@@ -390,6 +390,12 @@ func (st *relocSymState) relocsym(s loader.Sym, P []byte) {\n \t\t\to = ldr.SymValue(rs) + r.Add() - int64(ldr.SymSect(rs).Vaddr)\n \t\tcase objabi.R_WEAKADDROFF, objabi.R_METHODOFF:\n \t\t\tif !ldr.AttrReachable(rs) {\n+\t\t\t\tif...
2020-09-24T20:54:31
ollama/ollama
325d74985b9f31917ead1585ea22389a39b280b5
d9cd3d9667d83f68040378dc2834a49962e08244
Fix CPU performance on hyperthreaded systems The default thread count logic was broken and resulted in 2x the number of threads as it should on a hyperthreading CPU resulting in thrashing and poor performance.
[ { "path": "llm/ext_server.go", "patch": "@@ -37,7 +37,6 @@ import (\n \t\"fmt\"\n \t\"log\"\n \t\"os\"\n-\t\"runtime\"\n \t\"strings\"\n \t\"sync\"\n \t\"time\"\n@@ -185,11 +184,7 @@ func newExtServer(server extServer, model string, adapters, projectors []string,\n \t\tsparams.mmproj = nil\n \t}\n \n-\tif o...
2023-12-22T00:23:36
ggml-org/llama.cpp
59377a6c870be95e4c71715933e4e9ada71b8356
1239267cc4e5a1c9fc6546825eefe13c856e7458
ggml-backend: fix async set/get fallback sync (#19179)
[ { "path": "ggml/src/ggml-backend.cpp", "patch": "@@ -258,6 +258,7 @@ void ggml_backend_tensor_set_async(ggml_backend_t backend, struct ggml_tensor *\n GGML_ASSERT(offset + size <= ggml_nbytes(tensor) && \"tensor write out of bounds\");\n \n if (backend->iface.set_tensor_async == NULL) {\n+ gg...
2026-02-02T09:00:05
huggingface/transformers
5ad960f1f4f77f436ddf3de3692d09949a27c2df
65ea1904ff5dcd53f9ee97c40d2552398cecb572
Add Watermarking LogitsProcessor and WatermarkDetector (#29676) * add watermarking processor * remove the other hashing (context width=1 always) * make style * Update src/transformers/generation/logits_process.py Co-authored-by: Joao Gante <joaofranciscocardosogante@gmail.com> * Update src/transformers/...
[ { "path": "docs/source/en/generation_strategies.md", "patch": "@@ -173,6 +173,55 @@ your screen, one word at a time:\n An increasing sequence: one, two, three, four, five, six, seven, eight, nine, ten, eleven,\n ```\n \n+\n+## Watermarking\n+\n+The `generate()` supports watermarking the generated text by ra...
2024-05-14T08:31:39
nodejs/node
36732084db9d0ff59b6ce31e839450cd91a156be
e3503aca086f3f2f9587004ac594ed83a57f7443
util,assert: expose util.isDeepStrictEqual() Provide `util.isDeepStrictEqual()` that works like `assert.deepStrictEqual()` but returns a boolean rather than throwing an error. Several userland modules have needed this functionality and implemented it independently. This functionality already exists in Node.js core, s...
[ { "path": "doc/api/assert.md", "patch": "@@ -134,7 +134,7 @@ changes:\n * `expected` {any}\n * `message` {any}\n \n-Similar to `assert.deepEqual()` with the following exceptions:\n+Identical to [`assert.deepEqual()`][] with the following exceptions:\n \n 1. Primitive values besides `NaN` are compared using ...
2017-10-06T04:27:46
golang/go
0163bdae685c1b060f8108ac5af13ea6374555b1
815a5e29f434281c9ae3740ad43aadd4464ae15e
crypto/tls: fix TestLinkerGC test A test that checks if "tls.(*Conn)" appears in any symbol's name. tls.Conn is a type, so the string "tls.(*Conn)" can only appear in the name of a method of Conn. But the test code doesn't use any of the methods. Not sure why this needs to be live. In particular, the linker is now abl...
[ { "path": "src/crypto/tls/link_test.go", "patch": "@@ -41,19 +41,6 @@ func main() {}\n \t\t\t\t\"type.crypto/tls.serverHandshakeState\",\n \t\t\t},\n \t\t},\n-\t\t{\n-\t\t\tname: \"only_conn\",\n-\t\t\tprogram: `package main\n-import \"crypto/tls\"\n-var c = new(tls.Conn)\n-func main() {}\n-`,\n-\t\t\twant:...
2020-09-28T21:43:44
ggml-org/llama.cpp
b4d05a3d2fc7820444ca618570a3ac76cc12fe83
2dc3ce2166a6c3b149402af60c15f4d80b472a6c
spec : various improvements ton ngram-map + docs (#19253) * spec: ngram-map and reasoning chats * spec: add t_begin and t_accept * ngram-map : add internal hash map * docs : update ngram-map, add ngram-mod * docs : fix ngram-map-k * docs : differences between implementations
[ { "path": "common/ngram-map.cpp", "patch": "@@ -7,6 +7,18 @@\n #include <cstdio>\n #include <sstream>\n \n+// prime number used for LCG hash function (32 bit), it is near (sqrt(5) - 1)/2 * 2^32.\n+#define LCG_FACTOR 2654435761UL\n+\n+// Compute the LCG hash of a n-gram of size len at offset start.\n+static ...
2026-02-02T06:26:58
huggingface/transformers
c02d302e6bb92382fda8aabf50c6bc0b738405bd
449894d2e599f78d71a31fa8874ea69d86cedaf3
Fix cache type in Idefics2 (#30729) standardize cache in idefics2
[ { "path": "src/transformers/models/idefics2/modeling_idefics2.py", "patch": "@@ -1591,9 +1591,10 @@ def forward(\n raise ValueError(\"You have to specify either input_ids or inputs_embeds\")\n \n past_seen_tokens = 0\n- if use_cache:\n- if not isinstance(past_key_values...
2024-05-14T08:30:53
rust-lang/rust
596c14ade49d94bd1cfaa26aa34f7ef016f632d6
8c392966a013fd8a09e6b78b3c8d6e442bc278e1
fix `RangeBounds::is_empty` documentation One-sided ranges are never empty
[ { "path": "library/core/src/ops/range.rs", "patch": "@@ -827,7 +827,7 @@ pub trait RangeBounds<T: ?Sized> {\n }\n \n /// Returns `true` if the range contains no items.\n- /// One-sided ranges (`RangeFrom`, etc) always return `true`.\n+ /// One-sided ranges (`RangeFrom`, etc) always return `fal...
2025-03-01T21:38:07
ggml-org/llama.cpp
8a98ba4582ea961f06d350e60cf3572015489745
2634ed207a17db1a54bd8df0555bd8499a6ab691
nix: fix allowUnfreePredicate for packages with multiple licenses (#19237) The allowUnfreePredicate in pkgsCuda was wrapping p.meta.license in a list unconditionally. This fails when meta.license is already a list of licenses, as it creates a nested list and then tries to access .free and .shortName on the inner list....
[ { "path": ".devops/nix/nixpkgs-instances.nix", "patch": "@@ -4,7 +4,7 @@\n # the module `{ pkgs ... }: { /* config */ }` implicitly uses\n # `_module.args.pkgs` (defined in this case by flake-parts).\n perSystem =\n- { system, ... }:\n+ { lib, system, ... }:\n {\n _module.args = {\n ...
2026-02-01T20:10:48
nodejs/node
e3503aca086f3f2f9587004ac594ed83a57f7443
e8de2bfd138c42f4632777c01d3ef79a03fb8333
src: destroy inspector agent before context The inspector_agent depends on the context still being accessible during the destructor execution. PR-URL: https://github.com/nodejs/node/pull/16472 Fixes: https://github.com/nodejs/node/issues/15558 Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl> Reviewed-By: Franziska Hin...
[ { "path": "src/env-inl.h", "patch": "@@ -306,7 +306,7 @@ inline Environment::Environment(IsolateData* isolate_data,\n emit_napi_warning_(true),\n makecallback_cntr_(0),\n #if HAVE_INSPECTOR\n- inspector_agent_(this),\n+ inspector_agent_(new inspector::Agent(this)),\n #endif\n han...
2017-10-25T05:24:28
huggingface/transformers
449894d2e599f78d71a31fa8874ea69d86cedaf3
37bba2a32d2742a10216ffd925bb8f145a732ce1
Fix OWLv2 Doc (#30794) fix: owlv2 doc
[ { "path": "docs/source/en/model_doc/owlv2.md", "patch": "@@ -64,8 +64,8 @@ OWLv2 is, just like its predecessor [OWL-ViT](owlvit), a zero-shot text-conditio\n >>> for box, score, label in zip(boxes, scores, labels):\n ... box = [round(i, 2) for i in box.tolist()]\n ... print(f\"Detected {text[label]}...
2024-05-14T06:36:11
ollama/ollama
63aac0edc59dd368050cb12a213af482dec71875
1ca484f67e6f607114496211004942013e5595eb
fix(test): use real version string for comparison
[ { "path": "server/routes_test.go", "patch": "@@ -16,6 +16,7 @@ import (\n \n \t\"github.com/jmorganca/ollama/api\"\n \t\"github.com/jmorganca/ollama/parser\"\n+\t\"github.com/jmorganca/ollama/version\"\n )\n \n func setupServer(t *testing.T) (*Server, error) {\n@@ -71,7 +72,7 @@ func Test_Routes(t *testing....
2023-12-19T23:02:37
ggml-org/llama.cpp
41ea26144e55d23f37bb765f88c07588d786567f
89f10baad5a1809055d71110dff60e55561b9c62
nix: fix nix develop .#python-scripts (#19218) Without this I get: > * Getting build dependencies for wheel... > * Building wheel... > Successfully built gguf-0.17.1-py3-none-any.whl > Finished creating a wheel... > Finished executing pypaBuildPhase > Running phase: pythonRuntimeDepsCheckHook > Executing pythonRuntim...
[ { "path": ".devops/nix/package-gguf-py.nix", "patch": "@@ -3,6 +3,7 @@\n llamaVersion,\n numpy,\n tqdm,\n+ requests,\n sentencepiece,\n pyyaml,\n poetry-core,\n@@ -20,6 +21,7 @@ buildPythonPackage {\n tqdm\n sentencepiece\n pyyaml\n+ requests\n ];\n src = lib.cleanSource ../....
2026-01-31T16:01:46
nodejs/node
e8de2bfd138c42f4632777c01d3ef79a03fb8333
e714c42eabb265d7f8a230de595eef12fddd58c9
2017-10-25, Version 8.8.1 (Current) Notable changes: * net: - Fix timeout with null handle issue. This is a regression in Node 8.8.0. https://github.com/nodejs/node/pull/16489
[ { "path": "CHANGELOG.md", "patch": "@@ -27,7 +27,8 @@ release.\n </tr>\n <tr>\n <td valign=\"top\">\n-<b><a href=\"doc/changelogs/CHANGELOG_V8.md#8.8.0\">8.8.0</a></b><br/>\n+<b><a href=\"doc/changelogs/CHANGELOG_V8.md#8.8.1\">8.8.1</a></b><br/>\n+<a href=\"doc/changelogs/CHANGELOG_V8.md#8.8.0\">8.8.0</...
2017-10-25T19:10:15
huggingface/transformers
37bba2a32d2742a10216ffd925bb8f145a732ce1
539ed75d501f9af87b28dc0325b8d2f62b791cd9
CI: update to ROCm 6.0.2 and test MI300 (#30266) * update to ROCm 6.0.2 and test MI300 * add callers for mi300 * update dockerfile * fix trainer tests * remove apex * style * Update tests/trainer/test_trainer_seq2seq.py * Update tests/trainer/test_trainer_seq2seq.py * Update tests/trainer/test_...
[ { "path": ".github/workflows/self-push-amd-mi300-caller.yml", "patch": "@@ -0,0 +1,25 @@\n+name: Self-hosted runner (AMD mi300 CI caller)\n+\n+on:\n+ workflow_run:\n+ workflows: [\"Self-hosted runner (push-caller)\"]\n+ branches: [\"main\"]\n+ types: [completed]\n+ push:\n+ branches:\n+ -...
2024-05-13T16:14:36
ollama/ollama
6558f94ed022a0d8ef0c06afa2df58fb1c298676
54dbfa4c4a2c52dc0c2361e65090a0ede3339a63
Fix darwin intel build
[ { "path": "gpu/gpu_darwin.go", "patch": "@@ -4,6 +4,8 @@ package gpu\n \n import \"C\"\n import (\n+\t\"runtime\"\n+\n \t\"github.com/jmorganca/ollama/api\"\n )\n \n@@ -25,8 +27,12 @@ func GetGPUInfo() GpuInfo {\n }\n \n func NumGPU(numLayer, fileSizeBytes int64, opts api.Options) int {\n-\t// default to en...
2023-12-19T21:32:24
golang/go
815a5e29f434281c9ae3740ad43aadd4464ae15e
a28edbfca276307b228eb4b154bc2d137a3cba4a
cmd/go: fix doc math for build cache hashing The function takes five 24-bit chunks from the hash, resulting in 120 bits. When base-64 encoded, this results in a 20-byte output string, which is confirmed by "var dst [chunks * 4]byte". It seems like the documented math could have been written for a previous implementat...
[ { "path": "src/cmd/go/internal/work/buildid.go", "patch": "@@ -110,15 +110,15 @@ func contentID(buildID string) string {\n \n // hashToString converts the hash h to a string to be recorded\n // in package archives and binaries as part of the build ID.\n-// We use the first 96 bits of the hash and encode it ...
2020-09-26T20:00:37
ggml-org/llama.cpp
89f10baad5a1809055d71110dff60e55561b9c62
3dd95914d09b155eed84664b9abdbbffae238738
ggml-hexagon: flash-attention and reduce-sum optimizations (#19141) * wip * ggml-hexagon: add vectorized dot product function for FP32 and FP16 accumulation * ggml-hexagon: optimize dot product functions for FP16 and FP32 with new vectorized implementations * wip * ggml-hexagon: optimize hvx_vec_dump_f32_n and hvx...
[ { "path": "ggml/src/ggml-hexagon/CMakeLists.txt", "patch": "@@ -1,8 +1,20 @@\n file(TO_CMAKE_PATH \"${HEXAGON_SDK_ROOT}\" HEXAGON_SDK_ROOT)\n file(TO_CMAKE_PATH \"${HEXAGON_TOOLS_ROOT}\" HEXAGON_TOOLS_ROOT)\n \n-if (NOT IS_DIRECTORY \"${HEXAGON_SDK_ROOT}\" OR NOT IS_DIRECTORY \"${HEXAGON_TOOLS_ROOT}\")\n-...
2026-01-31T05:14:20
nodejs/node
e714c42eabb265d7f8a230de595eef12fddd58c9
a3998817739e22214fe65c3d122deac43bb47b01
build: fix npm install with --shared The npm install rules had a hidden dependency on the `node` binary install rule creating the `$PREFIX/bin` directory. Because with `./configure --shared` no binary is created, the rule subsequently failed. Fix that by creating the directory before creating the symlinks to the npm...
[ { "path": "tools/install.py", "patch": "@@ -33,6 +33,7 @@ def try_unlink(path):\n def try_symlink(source_path, link_path):\n print 'symlinking %s -> %s' % (source_path, link_path)\n try_unlink(link_path)\n+ try_mkdir_r(os.path.dirname(link_path))\n os.symlink(source_path, link_path)\n \n def try_mkdi...
2017-10-24T11:01:50
huggingface/transformers
0f8fefd4818a575dd749ea919abf4543afffbabf
82c1625ec3cb2c54f8a40cb00ca03eade3206e1b
Deprecate models script (#30184) * Add utility for finding candidate models for deprecation * Update model init * Make into configurable script * Fix path * Add sorting of base object alphabetically * Tidy * Refactor __init__ alpha ordering * Update script with logging * fix import * Fix log...
[ { "path": "utils/deprecate_models.py", "patch": "@@ -0,0 +1,357 @@\n+\"\"\"\n+Script which deprecates a list of given models\n+\n+Example usage:\n+python utils/deprecate_models.py --models bert distilbert\n+\"\"\"\n+\n+import argparse\n+import os\n+from collections import defaultdict\n+from pathlib import P...
2024-05-13T15:30:55
rust-lang/rust
28a5eeb157329fd45c31d7c5567bc48ecfdd4aa9
497f14013c31211f4d5d4f7c61813babad94679e
Fix `test_keyword_highlighting` test
[ { "path": "src/tools/rust-analyzer/crates/ide/src/syntax_highlighting/test_data/highlight_keywords_2015.html", "patch": "@@ -54,21 +54,21 @@\n <span class=\"brace\">}</span>\n \n <span class=\"keyword\">macro_rules</span><span class=\"macro_bang\">!</span> <span class=\"macro declaration public\">void</span...
2025-03-01T16:49:51
ollama/ollama
f8ef4439e9673c7df2314fafb5975aeab856c51f
d4cd6957598ba6a3a1bb4e2660ee24b82e2541da
Use build tags to generate accelerated binaries for CUDA and ROCm on Linux. The build tags rocm or cuda must be specified to both go generate and go build. ROCm builds should have both ROCM_PATH set (and the ROCM SDK present) as well as CLBlast installed (for GGML) and CLBlast_DIR set in the environment to the CLBlast...
[ { "path": "Dockerfile", "patch": "@@ -11,8 +11,8 @@ RUN mkdir -p /usr/local && tar xz -C /usr/local </tmp/go1.21.3.tar.gz\n COPY . .\n ENV GOARCH=$TARGETARCH\n ENV GOFLAGS=$GOFLAGS\n-RUN /usr/local/go/bin/go generate ./... \\\n- && /usr/local/go/bin/go build .\n+RUN /usr/local/go/bin/go generate -tags cu...
2023-10-17T00:41:40
ggml-org/llama.cpp
3dd95914d09b155eed84664b9abdbbffae238738
ec6c7421e4f3c0e1bdb9fdc6ee9670e14a490854
quantize: add option --tensor-type-file to llama-quantize (#18572) * add option --tensor-type-file to llama-quantize, but it raises an error. * add error message when file not found * quantize: update help menu, fix CI Signed-off-by: Aaron Teo <aaron.teo1@ibm.com> --------- Signed-off-by: Aaron Teo <aaron.teo1@ib...
[ { "path": "tools/quantize/quantize.cpp", "patch": "@@ -119,7 +119,7 @@ static bool try_parse_ftype(const std::string & ftype_str_in, llama_ftype & ftyp\n [[noreturn]]\n static void usage(const char * executable) {\n printf(\"usage: %s [--help] [--allow-requantize] [--leave-output-tensor] [--pure] [--ima...
2026-01-31T03:39:21
golang/go
a28edbfca276307b228eb4b154bc2d137a3cba4a
79e681d2a291142aa0ac8297229e182b2d1a78ac
encoding/asn1: error instead of panic on invalid value to Unmarshal Changes Unmarshal to return an error, instead of panicking when its value is nil or not a pointer. This change matches the behavior of other encoding packages like json. Fixes #41509. Change-Id: I92c3af3a960144566e4c2b55d00c3a6fe477c8d5 GitHub-Last...
[ { "path": "src/encoding/asn1/asn1.go", "patch": "@@ -1035,7 +1035,8 @@ func setDefaultValue(v reflect.Value, params fieldParameters) (ok bool) {\n // Unmarshal parses the DER-encoded ASN.1 data structure b\n // and uses the reflect package to fill in an arbitrary value pointed at by val.\n // Because Unmars...
2020-09-29T10:03:38
nodejs/node
cecbb595d5c4ea97264092abf2ade16b54a945fd
a78327f48be4266e250ad4c2170a0f263b47bc5f
net: fix timeout with null handle This commit handles the case where _onTimeout is called with a null handle. Refs: https://github.com/nodejs/node/pull/15791 Fixes: https://github.com/nodejs/node/issues/16484 PR-URL: https://github.com/nodejs/node/pull/16489 Reviewed-By: Anna Henningsen <anna@addaleax.net> Reviewed-B...
[ { "path": "lib/net.js", "patch": "@@ -397,13 +397,15 @@ Socket.prototype.setTimeout = function(msecs, callback) {\n \n \n Socket.prototype._onTimeout = function() {\n- // `.prevWriteQueueSize` !== `.updateWriteQueueSize()` means there is\n- // an active write in progress, so we suppress the timeout.\n- c...
2017-10-25T13:26:20
ollama/ollama
c063ee4af0de13287569fb61e2f9cc2ea1f8cd22
d99fa6ce0ad168ac488c5a741ca178caa12dc810
update runner submodule to fix hipblas build
[ { "path": "llm/llama.cpp/gguf", "patch": "@@ -1 +1 @@\n-Subproject commit b9e74f9bca5fdf7d0a22ed25e7a9626335fdfa48\n+Subproject commit 3c04bf6da89eaf4c7d317e0518f0687dfcbf2de7", "additions": 1, "deletions": 1, "language": "Unknown" } ]
2023-12-18T20:41:13
ggml-org/llama.cpp
1488339138d609139c4400d1b80f8a5b1a9a203c
49277958109f738eb5377242e46734788ba21be8
lookup, lookahead: fix crash when n_ctx not specified (#18729) * lookup, lookahead: fix crash when n_ctx not specified Since PR #16653 (Dec 15, 2025), the default n_ctx is 0 to enable automatic GPU memory fitting. This causes llama-lookup and llama-lookahead to crash when run without explicit -c flag: GGML_ASSER...
[ { "path": "examples/lookahead/lookahead.cpp", "patch": "@@ -50,6 +50,12 @@ int main(int argc, char ** argv) {\n const int N = 5; // n-gram size\n const int G = 15; // max verification n-grams\n \n+ // lookahead requires W + G + 1 sequences for parallel Jacobi decoding\n+ params.n_parallel = W...
2026-01-30T20:10:24
huggingface/transformers
94306352f489c7c2a8dc18af89e2efe0a76a5159
de2f722172089473a0d1ff0c037cd6b29460493f
Port IDEFICS to tensorflow (#26870) * Initial commit * Just a copy of modeling_idefics.py that will be ported to TF * - Prepend TF to the name of all classes - Convert pytorch ops to TF (not all operations are converted yet) * Add TF imports * Add autotranslated files * Add TF classes to model_tf_auto....
[ { "path": "docs/source/en/index.md", "patch": "@@ -160,7 +160,7 @@ Flax), PyTorch, and/or TensorFlow.\n | [HerBERT](model_doc/herbert) | ✅ | ✅ | ✅ |\n | [Hubert](model_doc/hubert) ...
2024-05-13T14:59:46
golang/go
79e681d2a291142aa0ac8297229e182b2d1a78ac
9a7a981ab796038e9ddb148dabd97067d9cbbb01
crypto/tls: make config.Clone return nil if the source is nil Fixes #40565 Change-Id: I13a67be193f8cd68df02b8729529e627a73d364b GitHub-Last-Rev: b03d2c04fd88db909b40dfd7bd08fe13d8994ab9 GitHub-Pull-Request: golang/go#40566 Reviewed-on: https://go-review.googlesource.com/c/go/+/246637 Run-TryBot: Emmanuel Odeke <emm.o...
[ { "path": "src/crypto/tls/common.go", "patch": "@@ -727,9 +727,12 @@ func (c *Config) ticketKeyFromBytes(b [32]byte) (key ticketKey) {\n // ticket, and the lifetime we set for tickets we send.\n const maxSessionTicketLifetime = 7 * 24 * time.Hour\n \n-// Clone returns a shallow clone of c. It is safe to clo...
2020-09-29T09:05:41
nodejs/node
a78327f48be4266e250ad4c2170a0f263b47bc5f
d2e44d5e7fcc93c615ac5cef794bdfbcd32c3fc0
crypto: migrate setEngine to internal/errors PR-URL: https://github.com/nodejs/node/pull/16429 Reviewed-By: Luigi Pinca <luigipinca@gmail.com> Reviewed-By: Michaël Zasso <targos@protonmail.com> Reviewed-By: Ali Ijaz Sheikh <ofrobots@google.com> Reviewed-By: Joyee Cheung <joyeec9h3@gmail.com>
[ { "path": "doc/api/errors.md", "patch": "@@ -637,6 +637,12 @@ Used when the native call from `process.cpuUsage` cannot be processed properly.\n Used when an invalid value for the `format` argument has been passed to the\n `crypto.ECDH()` class `getPublicKey()` method.\n \n+<a id=\"ERR_CRYPTO_ENGINE_UNKNOWN\...
2017-10-24T06:06:53
ggml-org/llama.cpp
49277958109f738eb5377242e46734788ba21be8
971facc38e2544fcf2cc09368de5d1a68e33c10f
ngram-mod : fix build [no ci] (#19216)
[ { "path": "common/ngram-mod.h", "patch": "@@ -2,6 +2,7 @@\n \n #include <cstdint>\n #include <vector>\n+#include <cstddef>\n \n //\n // common_ngram_mod", "additions": 1, "deletions": 0, "language": "C/C++ Header" } ]
2026-01-30T19:27:27
huggingface/transformers
69d9bca55af0f25a3b1a2cde5df3b3c8d42cb3ad
f4dc26d46687f5f4baf3fe64a1d87cafefbeec53
enable Pipeline to get device from model (#30534) * check model.device * fix * style fix * move model device * remove print * add comment * fix * add unit test * optimize * change test names and add more cases * Update tests/pipelines/test_pipelines_common.py Co-authored-by: amyerober...
[ { "path": "src/transformers/pipelines/base.py", "patch": "@@ -845,6 +845,8 @@ def __init__(\n device = -1\n \n if is_torch_available() and self.framework == \"pt\":\n+ if device == -1 and self.model.device is not None:\n+ device = self.model.device\n ...
2024-05-13T14:00:39
ollama/ollama
4a1abfe4fa67baa821834abc990d2c79ee252469
bbd41494bf5624f12f983a55b424411e700498bd
fix tests
[ { "path": "cmd/cmd.go", "patch": "@@ -951,7 +951,8 @@ func generateInteractive(cmd *cobra.Command, opts generateOptions) error {\n \t\t\t\t\tcmd.SetContext(ctx)\n \t\t\t\t}\n \t\t\t\tif len(opts.Images) == 0 {\n-\t\t\t\t\tfmt.Println(\"This model requires you to add a jpeg, png, or svg image.\\n\")\n+\t\t\t...
2023-12-13T19:42:30
golang/go
ad0ab812f8b80416c92ed227974e3194e98f4cdc
c4971a14a7cac78849f4d0908e7140263129bdf7
cmd/compile: fix type checking of "make" arguments As part of type checking make's arguments, we were converting untyped float and complex constant arguments to integers. However, we were doing this without concern for whether the argument was a declared constant. Thus a call like "make([]T, n)" could change n from an...
[ { "path": "src/cmd/compile/internal/gc/typecheck.go", "patch": "@@ -1770,7 +1770,7 @@ func typecheck1(n *Node, top int) (res *Node) {\n \t\t\t\tn.Type = nil\n \t\t\t\treturn n\n \t\t\t}\n-\t\t\tif !checkmake(t, \"len\", l) || r != nil && !checkmake(t, \"cap\", r) {\n+\t\t\tif !checkmake(t, \"len\", &l) || r...
2020-09-28T19:19:56
nodejs/node
8a00f1d129ac5fcbbce8e115a2576e0ea080e4ed
8172f4547e7e2a7e1b7363e1cb9d7431fad241ff
test: call toLowerCase on the resolved module The commit updates test-require-resolve.js to call toLowerCase on the resolved module instead of the path. Currently this test will fail if the path to where node exists contains uppercase letters. For example: ``` $ out/Release/node test/parallel/test-require-resolve.js ...
[ { "path": "test/parallel/test-require-resolve.js", "patch": "@@ -32,7 +32,7 @@ assert.strictEqual(\n require.resolve(fixtures.path('a')).toLowerCase());\n assert.strictEqual(\n fixtures.path('nested-index', 'one', 'index.js').toLowerCase(),\n- require.resolve(fixtures.path('nested-index', 'one').toLowe...
2017-10-25T10:36:24
ggml-org/llama.cpp
971facc38e2544fcf2cc09368de5d1a68e33c10f
d9a2a4bcaa071d730bb1ab4fb411a9c93b50dd13
opencl: add optimized q8_0 mm kernel for adreno (#18871) * Add Q8_0 OpenCL kernel Co-authored-by: yunjie <yunjie@qti.qualcomm.com> * opencl: fix build for non-adreno * opencl: refactor q8_0 * opencl: enforce subgroup size of 64 for adreno for q8_0 * For A750 and older generations, subgroup size can be 64 or 128. ...
[ { "path": "ggml/src/ggml-opencl/CMakeLists.txt", "patch": "@@ -101,6 +101,8 @@ set(GGML_OPENCL_KERNELS\n mul_mm_f32_f32_l4_lm\n mul_mm_f16_f32_l4_lm\n mul_mm_q8_0_f32_l4_lm\n+ mul_mm_q8_0_f32_8x4\n+ gemv_noshuffle_general_q8_0_f32\n mul\n norm\n relu", "additions": 2, "...
2026-01-30T18:19:27
huggingface/transformers
f4dc26d46687f5f4baf3fe64a1d87cafefbeec53
f823fec53e6c18542a6064d28e4b4ada03e50c5b
Qwen: incorrect setup flag (#30776) qwen does not support the new cache classes
[ { "path": "src/transformers/models/qwen2/modeling_qwen2.py", "patch": "@@ -821,7 +821,6 @@ class Qwen2PreTrainedModel(PreTrainedModel):\n _skip_keys_device_placement = \"past_key_values\"\n _supports_flash_attn_2 = True\n _supports_sdpa = True\n- _supports_cache_class = True\n \n def _ini...
2024-05-13T13:12:58
ollama/ollama
0a9d34802375641d72279fd9870e2ddfd4120fbf
3144e2a4394a544375ac054375d838525f8096c7
Fix issues with `/set template` and `/set system` (#1486)
[ { "path": "README.md", "patch": "@@ -104,7 +104,7 @@ FROM llama2\n # set the temperature to 1 [higher is more creative, lower is more coherent]\n PARAMETER temperature 1\n \n-# set the system prompt\n+# set the system message\n SYSTEM \"\"\"\n You are Mario from Super Mario Bros. Answer as Mario, the assist...
2023-12-12T19:43:19
golang/go
f33263d11aa0f5e2668bfc6a0805e4edee17b03c
a424f6e45e29960c933a7ccc1cd8fc9bb2766f15
crypto/x509: hardcode RSA PSS parameters rather than generating them Rather than generating the three possible RSA PSS parameters each time they are needed just hardcode them and pick the required one based on the hash function. Fixes #41407 Change-Id: Id43bdaf40b3ca82c4c04c6588e3b643f63107657 Reviewed-on: https://g...
[ { "path": "src/crypto/x509/x509.go", "patch": "@@ -351,6 +351,19 @@ var signatureAlgorithmDetails = []struct {\n \t{PureEd25519, \"Ed25519\", oidSignatureEd25519, Ed25519, crypto.Hash(0) /* no pre-hashing */},\n }\n \n+// hashToPSSParameters contains the DER encoded RSA PSS parameters for the\n+// SHA256, S...
2020-09-28T15:59:13
ggml-org/llama.cpp
dfd6106c84045f819760c86cc21a27804d2faf32
bbada8bfb9bce17eafb0af5ded22bf96ba57f778
cuda : fix compile warnings (whisper/0)
[ { "path": "ggml/src/ggml-cuda/ggml-cuda.cu", "patch": "@@ -3910,14 +3910,14 @@ static void ggml_cuda_graph_evaluate_and_capture(ggml_backend_cuda_context * cud\n // Launch graph\n CUDA_CHECK(cudaGraphLaunch(graph->instance, cuda_ctx->stream()));\n #else\n+ GGML_UNUSED(graph_key);\n ...
2026-01-30T13:56:15
nodejs/node
3621889c800707f3ca720c8de5f05a798a1fe5a8
73533a193209032e192f6250b4d194a6382837e2
doc: improve http2 documentation Provide section headings for server-side and client side examples. Add error handling and TLS to server-side example, following example of `https`. Add error handling, TLS, more efficient Buffer usage, and header printing to client example. PR-URL: https://github.com/nodejs/node/pull/...
[ { "path": "doc/api/http2.md", "patch": "@@ -16,14 +16,25 @@ support for HTTP/2 protocol features. It is specifically *not* designed for\n compatibility with the existing [HTTP/1][] module API. However,\n the [Compatibility API][] is.\n \n+The `http2` Core API is much more symmetric between client and server...
2017-10-21T19:04:35
ollama/ollama
5314fc9b632e520415114b4bfd4e53bab263467c
a36b5fef3b83fc622b3720d7b3a71f1261ba193e
Fix Readme "Database -> MindsDB" link (#1479)
[ { "path": "README.md", "patch": "@@ -256,7 +256,7 @@ See the [API documentation](./docs/api.md) for all endpoints.\n \n ### Database\n \n-- [MindsDB] (https://github.com/mindsdb/mindsdb/blob/staging/mindsdb/integrations/handlers/ollama_handler/README.md)\n+- [MindsDB](https://github.com/mindsdb/mindsdb/blob...
2023-12-12T15:26:13
huggingface/transformers
f823fec53e6c18542a6064d28e4b4ada03e50c5b
a0779b9e19093dc0371abbf516030491eec3d86c
Generation / FIX: Fix multi-device generation (#30746) * attempt to fix multi-device generation * fix * final fix * final fix * fix * fix * fix * fix * add joao suggestion * fix
[ { "path": "src/transformers/generation/utils.py", "patch": "@@ -476,6 +476,7 @@ def _prepare_attention_mask_for_generation(\n )\n can_infer_attention_mask = is_pad_token_in_inputs * is_pad_token_not_equal_to_eos_token_id\n attention_mask_from_padding = inputs.ne(pad_token_id).long()\...
2024-05-13T12:35:45
golang/go
874b3132a84cf76da6a48978826c04c380a37a50
05b626e49075d3b9f2fcda65c7cc9054381da047
mime/multipart: return overflow errors in Reader.ReadForm Updates Reader.ReadForm to check for overflow errors that may result from a leeway addition of 10MiB to the input argument maxMemory. Fixes #40430 Change-Id: I510b8966c95c51d04695ba9d08fcfe005fd11a5d Reviewed-on: https://go-review.googlesource.com/c/go/+/2474...
[ { "path": "src/mime/multipart/formdata.go", "patch": "@@ -7,6 +7,7 @@ package multipart\n import (\n \t\"bytes\"\n \t\"errors\"\n+\t\"fmt\"\n \t\"io\"\n \t\"io/ioutil\"\n \t\"net/textproto\"\n@@ -41,6 +42,9 @@ func (r *Reader) readForm(maxMemory int64) (_ *Form, err error) {\n \n \t// Reserve an additional ...
2020-08-07T18:50:12
ggml-org/llama.cpp
dabaa2e77a9d94ee1466ce55f9ff47d5c1daae43
2e916f996a965a911eb79168b5fb959c1ea0c57b
spec : add ngram-mod (#19164) * spec : add ngram-mod * cont : simplify + keep track of occupancy * cont : cleanup * cont : move initialization to common/speculative * cont : cleanup * cont : cleanup * cont : fix
[ { "path": "common/CMakeLists.txt", "patch": "@@ -75,6 +75,8 @@ add_library(${TARGET} STATIC\n ngram-cache.h\n ngram-map.cpp\n ngram-map.h\n+ ngram-mod.cpp\n+ ngram-mod.h\n peg-parser.cpp\n peg-parser.h\n preset.cpp", "additions": 2, "deletions": 0, "language": "Plai...
2026-01-30T16:21:48
nodejs/node
73533a193209032e192f6250b4d194a6382837e2
ed116dc3c65a6d568e2de7c09bd1c6f9db2c9b24
http2: do not allow socket manipulation Because of the specific serialization and processing requirements of HTTP/2, sockets should not be directly manipulated. This forbids any interactions with destroy, emit, end, pause, read, resume and write methods of the socket. It also redirects setTimeout to session instead of...
[ { "path": "doc/api/errors.md", "patch": "@@ -810,8 +810,8 @@ reached.\n <a id=\"ERR_HTTP2_NO_SOCKET_MANIPULATION\"></a>\n ### ERR_HTTP2_NO_SOCKET_MANIPULATION\n \n-Used when attempting to read, write, pause, and/or resume a socket attached to\n-an `Http2Session`.\n+Used when attempting to directly manipulat...
2017-10-20T01:32:20
ollama/ollama
7db5bcf73bf7026970e988f56126db8f370f1b11
fa2f095bd9f002c298042aed5fdc0e7ccd50a652
fix `go-staticcheck` warning
[ { "path": "server/routes.go", "patch": "@@ -860,7 +860,7 @@ func Serve(ln net.Listener, allowOrigins []string) error {\n \tif runtime.GOOS == \"linux\" {\n \t\t// check compatibility to log warnings\n \t\tif _, err := llm.CheckVRAM(); err != nil {\n-\t\t\tlog.Printf(err.Error())\n+\t\t\tlog.Print(err.Error(...
2023-12-10T16:44:27
huggingface/transformers
a0779b9e19093dc0371abbf516030491eec3d86c
453893ed154e5cd31ecf7d4b1b71189aeb29a7f0
Llama: fix custom 4D masks, v2 (#30348) * 4d mask fixes * Update custom 4D mask logic * test moved to mixin * extra tests 4d mask * upd 4d mask and StaticCache handling * added Mask4DTestHard to mistral tests * post-rebase fixes * test fixes for StaticCache * make fix-copies * upd 1 after #3...
[ { "path": "src/transformers/modeling_attn_mask_utils.py", "patch": "@@ -250,7 +250,7 @@ def _ignore_causal_mask_sdpa(\n allowing to dispatch to the flash attention kernel (that can otherwise not be used if a custom `attn_mask` is passed).\n \"\"\"\n \n- batch_size, query_length = inpu...
2024-05-13T11:46:06
golang/go
05b626e49075d3b9f2fcda65c7cc9054381da047
e572218d1273bf54bf8cafd39f93f22de196dd55
os: fix SyscallConn typos in the File.Fd comments This CL fixes two typos introduced in CL 256899. Change-Id: I47f0a3097deeeec8d6e9bbe7073fcf7a28c5dff9 Reviewed-on: https://go-review.googlesource.com/c/go/+/257997 Trust: Tobias Klauser <tobias.klauser@gmail.com> Trust: Emmanuel Odeke <emm.odeke@gmail.com> Run-TryBot:...
[ { "path": "src/os/file_plan9.go", "patch": "@@ -35,7 +35,7 @@ type file struct {\n // a finalizer might be run. On Unix systems this will cause the SetDeadline\n // methods to stop working.\n //\n-// As an alternative, see the f.SyscallCon method.\n+// As an alternative, see the f.SyscallConn method.\n func...
2020-09-28T07:46:42
ggml-org/llama.cpp
2e916f996a965a911eb79168b5fb959c1ea0c57b
f3bc98890c2b07f66c0ee20cd539abbfe9ea7639
jinja : add unordered_map include to value.h [no ci] (#19205) On macos Sequoia 15.7.3, x86_64, the build has recently started failing with ``` In file included from .../code/cpp/llama.cpp/common/jinja/string.cpp:2: .../code/cpp/llama.cpp/common/./jinja/value.h:478:10: error: no template named 'unordered_map' in nam...
[ { "path": "common/jinja/value.h", "patch": "@@ -12,6 +12,7 @@\n #include <set>\n #include <sstream>\n #include <string>\n+#include <unordered_map>\n #include <vector>\n \n namespace jinja {", "additions": 1, "deletions": 0, "language": "C/C++ Header" } ]
2026-01-30T15:09:44
nodejs/node
ed116dc3c65a6d568e2de7c09bd1c6f9db2c9b24
9c6f6b0633ef4009cc04cfff5efb29c23ef5fc2b
test: fix test for inherited properties on vm The known issue is fixed with https://github.com/nodejs/node/pull/16293. The text needs to call `Object.hasOwnProperty(this)` instead of `this.hasOwnProperty()`, otherwise `this` is from the wrong context is used. Add a second test case taken verbatim from issue https://...
[ { "path": "test/known_issues/test-vm-inherited_properties.js", "patch": "@@ -1,20 +0,0 @@\n-'use strict';\n-// Ref: https://github.com/nodejs/node/issues/5350\n-\n-require('../common');\n-const vm = require('vm');\n-const assert = require('assert');\n-\n-const base = {\n- propBase: 1\n-};\n-\n-const sandbo...
2017-10-23T14:06:35
ollama/ollama
fa2f095bd9f002c298042aed5fdc0e7ccd50a652
045b855db9dc4fa08738dd20ee2a7252c4680e6a
fix model name returned by `/api/generate` being different than the model name provided
[ { "path": "llm/llama.go", "patch": "@@ -545,15 +545,13 @@ type prediction struct {\n const maxBufferSize = 512 * format.KiloByte\n \n type PredictOpts struct {\n-\tModel string\n \tPrompt string\n \tFormat string\n \tCheckpointStart time.Time\n \tCheckpointLoaded time.Time\n ...
2023-12-10T16:42:15
golang/go
e572218d1273bf54bf8cafd39f93f22de196dd55
5755bad42adc23ad4a0c32149ac8cf78ece5d0b0
cmd/compile: fix escape reason for MAKESLICE with no cap When explaining why the slice from a make() call escapes for the -m -m message, we print "non-const size" if any one of Isconst(n.Left) and Isconst(n.Right) return false; but for OMAKESLICE nodes with no cap, n.Right is nil, so Isconst(n.Right, CTINT) will be al...
[ { "path": "src/cmd/compile/internal/gc/escape.go", "patch": "@@ -1053,7 +1053,7 @@ func (e *Escape) newLoc(n *Node, transient bool) *EscLocation {\n \n \t\tif mustHeapAlloc(n) {\n \t\t\twhy := \"too large for stack\"\n-\t\t\tif n.Op == OMAKESLICE && (!Isconst(n.Left, CTINT) || !Isconst(n.Right, CTINT)) {\n+...
2020-09-26T07:42:59
huggingface/transformers
e52741f60117acea1ab113db5e3b4e4245dd0d45
f63d822242e706e09308474b7b81c2f916d2d4a2
Support for Falcon2-11B (#30771) * remove unrelated changes * remove unrelated changes on phi and stable LM * add: Test for Falcon 10B * fix: formatting * fix: loading the falcon 10B in 8 bit precision using bitsanbytes. * fix: device placement * fix: broken tests. * fix: backwards compatibility f...
[ { "path": "src/transformers/models/falcon/configuration_falcon.py", "patch": "@@ -42,6 +42,9 @@ class FalconConfig(PretrainedConfig):\n Number of hidden layers in the Transformer decoder.\n num_attention_heads (`int`, *optional*, defaults to 71):\n Number of attention heads f...
2024-05-13T11:32:43
ggml-org/llama.cpp
05625031547d95f7335abd012cad4cd07986819b
83bcdf7217dc06ac67ff5f7322bdd89f46664c04
convert : add missing return statement for GraniteMoeModel (#19202) This commit adds a missing return statement to the GraniteMoeModel class to fix an issue in the model conversion process. Resolves: https://github.com/ggml-org/llama.cpp/issues/19201
[ { "path": "convert_hf_to_gguf.py", "patch": "@@ -8806,6 +8806,7 @@ def modify_tensors(self, data_torch: Tensor, name: str, bid: int | None) -> Iter\n gate, up = data_torch.split(ffn_dim, dim=-2)\n yield from ModelBase.modify_tensors(self, gate, self.format_tensor_name(gguf.MODEL_TENS...
2026-01-30T10:12:53
nodejs/node
9c6f6b0633ef4009cc04cfff5efb29c23ef5fc2b
4997894390f9e4e3dc078da6aba2ca52ad572055
module: support custom paths to require.resolve() This commit allows custom lookup paths to be passed to require.resolve(). It also adds require.resolve.paths() which retrieves the default resolution paths. Fixes: https://github.com/nodejs/node/issues/5963 Fixes: https://github.com/nodejs/node/issues/16389 PR-URL: ht...
[ { "path": "doc/api/modules.md", "patch": "@@ -598,14 +598,36 @@ filename scales linearly with the number of registered extensions.\n In other words, adding extensions slows down the module loader and\n should be discouraged.\n \n-#### require.resolve()\n+#### require.resolve(request[, options])\n <!-- YAML\...
2016-04-25T16:19:28
ggml-org/llama.cpp
d284baf1b5969b74177e705ee6bca63c1bbfb4b7
bd90fc74c3fecd18f36e26a91b3c3282578bf680
Fix typos in SYCL documentation (#19162) * Fix typos in SYCL documentation * Update SYCL.md * Update SYCL.md * Update SYCL.md * Update docs/backend/SYCL.md Co-authored-by: Neo Zhang Jianyu <jianyu.zhang@intel.com> * Update SYCL.md --------- Co-authored-by: Neo Zhang Jianyu <jianyu.zhang@intel.com>
[ { "path": "docs/backend/SYCL.md", "patch": "@@ -35,9 +35,9 @@ The following releases are verified and recommended:\n \n |Commit ID|Tag|Release|Verified Platform| Update date|\n |-|-|-|-|-|\n-|24e86cae7219b0f3ede1d5abdf5bf3ad515cccb8|b5377 |[llama-b5377-bin-win-sycl-x64.zip](https://github.com/ggml-org/llam...
2026-01-30T01:46:57
golang/go
5755bad42adc23ad4a0c32149ac8cf78ece5d0b0
8ab020adb27089fa207d015f2f69600ef3d1d307
os: remove ENOTSUP special case in Getwd on darwin ENOTSUP was used as a signaling error in the custom implementation of syscall.Getwd to fall back to the slow algorithm. Since CL 257637 Getwd directly calls the respective function from libSystem.dylib which can no longer return ENOTSUP. Change-Id: I8e65e42b3ea069bf7...
[ { "path": "src/os/getwd.go", "patch": "@@ -15,10 +15,6 @@ var getwdCache struct {\n \tdir string\n }\n \n-// useSyscallwd determines whether to use the return value of\n-// syscall.Getwd based on its error.\n-var useSyscallwd = func(error) bool { return true }\n-\n // Getwd returns a rooted path name corres...
2020-09-26T15:12:14