summaryrefslogtreecommitdiff
path: root/candle-core/examples
Commit message (Expand)AuthorAgeFilesLines
* Add a dequantize command to tensor-tools. (#1565)Laurent Mazare2024-01-111-1/+24
* Implement the module trait directly for QMatMul. (#1372)Laurent Mazare2023-11-251-6/+5
* Quantized version of mistral. (#1009)Laurent Mazare2023-09-301-9/+27
* Use yoke to provide a self-referential container for mmaped safetenso… (#939)Laurent Mazare2023-09-231-2/+1
* Use the proper block size for quantizing models. (#933)Laurent Mazare2023-09-221-2/+17
* T5 quantized example (#922)Laurent Mazare2023-09-211-0/+53
* Add a custom softmax implementation. (#744)Laurent Mazare2023-09-051-166/+0
* Dilated convolutions (#657)Laurent Mazare2023-08-293-6/+6
* Llama quantization. (#625)Laurent Mazare2023-08-271-15/+75
* Add the quantize command. (#624)Laurent Mazare2023-08-271-1/+75
* More pickle support. (#588)Laurent Mazare2023-08-241-1/+1
* Add to the cuda example a reproduction of the issue. (#579)Laurent Mazare2023-08-241-2/+11
* Add a test for conv2d with padding + bugfix the random number generation on c...Laurent Mazare2023-08-241-0/+3
* Add some group parameter to convolutions. (#566)Laurent Mazare2023-08-233-4/+4
* Handle GGUF files in tensor-tools. (#558)Laurent Mazare2023-08-231-1/+20
* Small tweaks to tensor-tools. (#517)Laurent Mazare2023-08-191-9/+15
* Retrieve tensor data from PyTorch files. (#516)Laurent Mazare2023-08-191-5/+7
* Retrieve more information from PyTorch checkpoints. (#515)Laurent Mazare2023-08-191-3/+9
* Add ggml support to tensor-tools (#512)Laurent Mazare2023-08-191-15/+59
* Preliminary support for importing PyTorch weights. (#511)Laurent Mazare2023-08-191-0/+16
* Add the tensor-tools binary. (#510)Laurent Mazare2023-08-191-0/+72
* Tensor -> QTensor conversion (#496)Laurent Mazare2023-08-181-1/+1
* AVX version of the vecdot for q4_0. (#474)Laurent Mazare2023-08-171-0/+24
* Cudnn support (#445)Laurent Mazare2023-08-141-5/+4
* Add a softmax bench. (#433)Laurent Mazare2023-08-131-1/+29
* Add a matmul benchmark. (#429)Laurent Mazare2023-08-131-0/+19
* More accelerate optimizations (#427)Laurent Mazare2023-08-132-0/+6
* Small example for benchmarking some cpu ops (#394)Laurent Mazare2023-08-102-24/+95
* Add a conv1d benchmark based on the whisper sizes. (#377)Laurent Mazare2023-08-091-0/+24
* Add some conv1d test + bugfix using padding. (#349)Laurent Mazare2023-08-081-20/+6
* Support the Accelerate BLAS on macOS. (#325)Laurent Mazare2023-08-051-0/+3
* Rename the candle crate to candle-core (#301)Laurent Mazare2023-08-023-3/+3
* Simplify Tensor::randn. (#255)Laurent Mazare2023-07-271-0/+5
* Simplify the parameters used by sum and sum_keepdim. (#165)Laurent Mazare2023-07-142-6/+6
* Use the same default as pytorch for sum. (#164)Laurent Mazare2023-07-132-10/+10
* Sketch a fast cuda kernel for reduce-sum. (#109)Laurent Mazare2023-07-081-0/+15
* Add some very simple sum benchmark. (#108)Laurent Mazare2023-07-082-34/+51
* Add mkl support for matrix multiply. (#86)Laurent Mazare2023-07-062-0/+6
* Move llama in a cargo-examples directory.laurent2023-07-034-912/+0
* Adding a bit more docs around safety.Nicolas Patry2023-07-031-1/+1
* Move more safetensors bits to the shared module.laurent2023-07-031-16/+8
* Move some safetensors bits in the candle-core crate.laurent2023-07-031-31/+2
* Add a flag for custom prompt.laurent2023-07-011-2/+7
* Early conversion for the llama weights.laurent2023-06-302-45/+19
* Add a const to easily tweak the dtype used for llama internal computations.laurent2023-06-301-4/+8
* Tweak the kv-cache flag.laurent2023-06-291-4/+4
* Add a flag.laurent2023-06-291-6/+11
* Enable the KV cache after fixing the caching length and the rope bits.laurent2023-06-291-14/+21
* Only narrow when needed + deactivate the kv cache.laurent2023-06-291-2/+6
* Add some KV cache to llama.laurent2023-06-291-36/+72
generated by cgit v1.2.3 (git 2.39.1) at 2025-04-05 17:29:17 +0000