-
Notifications
You must be signed in to change notification settings - Fork 931
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Add the F8E4M3 dtype for CUDA and CPU #2546
Commits on May 15, 2024
-
Mistral.rs Squash Changes (#4)
* Offset it * Freeze * Offset it * Offset it * Try out vllm impl again * Try out vllm impl again * Try out vllm impl again * Try out vllm impl again * Try out vllm impl again * Try out vllm impl again * Try out vllm impl again * Try out vllm impl again * Try out vllm impl again * Try out vllm impl again * Try out vllm impl again * Try out vllm impl again * Try out vllm impl again * Remove debugs * Polish it up * Polish it up * Clippy * Remove test file * Add config for if neox * Fix bug * Fix bug * Cast cache type on rust side * Cast types * To dtype * Drop temp * Update casting * Update casting * Update casting * Create dtype in bf16 * Check type * Debug * Check dtype * Check dtype * Check dtype * Check dtype * Check dtype * Check dtype * Check dtype * Check dtype * Check dtype * Debug * Debug * Debug * Check old method * Check old method * Check old method * Check old method * Check old method * Check old method * Check old method * Check old method * Check old method * Check old method * Check old method * Check old method * Check old method * Check old method * Check old method * Check old method * Check old method * Check old method * Check old method * Check old method * Check old method * Use mistral slow rope impl * Use mistral slow rope impl * Use mistral slow rope impl * Use mistral slow rope impl * Use mistral slow rope impl * Use mistral slow rope impl * Use mistral slow rope impl * Use mistral slow rope impl * Use mistral slow rope impl * Use mistral slow rope impl * Use mistral slow rope impl * Use mistral slow rope impl * Reseting * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Remove debug * Debug * Debug * Remove debug * Remove debug * Debug * Remove debug * Debug * Remove debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Debug * Try to use 3dim rotemb fused * Try to use 3dim rotemb fused * Remove contig and debug * Check handling * Cleanup * Fix * Remove prints * Lower block dim * Use fused layernorm * Pass batch size * Simplify internal API * Simplify internal API * Try slow * Try candle layer norm * Try candle layer norm * Fix dep of candle layer norm * Reshape input for rank 2 * Reshape input for rank 2 * Fix ref * Code style * Make dep optional * Ensure contig * Ensure contig * Ensure contig * Debug contig dmmv error * Debug contig dmmv error * Debug contig dmmv error * Debug contig dmmv error * Try other method * Try other method * Try other method * Try other method * Try other method * Use typestate to optimize * Use typestate to optimize * Fixes * Fixes * Fixes * Fixes * Fixes * Debug via using slow rmsnorm * Debug via using slow rope * Remove debug * More debugging * Remove debug * Remove debug * Remove debug * Add better error enum * Fix diff marker * Fix some things * Fix some things * Fix some things * Fix dummy backends * Re add from storage noop * Fix removed kvconcat custom op * Fix erroneous feature gate * Complete metal backend refactoring * Check if calling * Check if calling * Update default for force dmmv * Load atomic * Debug * Use mmvq * Update * Add the empty functions * Add rope new_partial function * Make variant of qmatmul pub * Make variant of qmatmul pub * Add the varbuilder set_device function * Only link stdc++ if target has msvc * Only link stdc++ if target has msvc * Only link stdc++ if target has msvc * Only link stdc++ if target has msvc * Handle case of device mapping * Handle case of device mapping * Add getter * Fix * Fix * Support nvcc flags in flash attn * Support nvcc flags in flash attn * Support nvcc flags in flash attn * Support nvcc flags in flash attn * Support nvcc flags in flash attn * Fixes * Fixes * Fix the tests * Fix the tests
Configuration menu - View commit details
-
Copy full SHA for 83a9e88 - Browse repository at this point
Copy the full SHA 83a9e88View commit details -
Configuration menu - View commit details
-
Copy full SHA for 4e82fab - Browse repository at this point
Copy the full SHA 4e82fabView commit details
Commits on May 16, 2024
-
Configuration menu - View commit details
-
Copy full SHA for 37cafcc - Browse repository at this point
Copy the full SHA 37cafccView commit details -
Configuration menu - View commit details
-
Copy full SHA for 5892fac - Browse repository at this point
Copy the full SHA 5892facView commit details
Commits on May 18, 2024
-
Configuration menu - View commit details
-
Copy full SHA for 9b151f5 - Browse repository at this point
Copy the full SHA 9b151f5View commit details
Commits on May 19, 2024
-
* Support flash-attn in quantized phi3. (huggingface#2194) * Use flash-attn in gemma. (huggingface#2195) * Use flash-attn in gemma. * Fix flash-attn for head dim 256. * Remove candle-layer-norm --------- Co-authored-by: Laurent Mazare <[email protected]>
Configuration menu - View commit details
-
Copy full SHA for ea49ea2 - Browse repository at this point
Copy the full SHA ea49ea2View commit details -
Configuration menu - View commit details
-
Copy full SHA for 38f8d9e - Browse repository at this point
Copy the full SHA 38f8d9eView commit details
Commits on May 27, 2024
-
Configuration menu - View commit details
-
Copy full SHA for c10fc33 - Browse repository at this point
Copy the full SHA c10fc33View commit details
Commits on May 28, 2024
-
Configuration menu - View commit details
-
Copy full SHA for 527ebcc - Browse repository at this point
Copy the full SHA 527ebccView commit details
Commits on May 29, 2024
-
Configuration menu - View commit details
-
Copy full SHA for bfc197b - Browse repository at this point
Copy the full SHA bfc197bView commit details
Commits on May 30, 2024
-
Configuration menu - View commit details
-
Copy full SHA for 0c2ac76 - Browse repository at this point
Copy the full SHA 0c2ac76View commit details
Commits on Jun 1, 2024
-
Configuration menu - View commit details
-
Copy full SHA for cb3dbc2 - Browse repository at this point
Copy the full SHA cb3dbc2View commit details
Commits on Jun 3, 2024
-
Configuration menu - View commit details
-
Copy full SHA for faa9435 - Browse repository at this point
Copy the full SHA faa9435View commit details -
Configuration menu - View commit details
-
Copy full SHA for 462d948 - Browse repository at this point
Copy the full SHA 462d948View commit details
Commits on Jun 4, 2024
-
Configuration menu - View commit details
-
Copy full SHA for 5c06acd - Browse repository at this point
Copy the full SHA 5c06acdView commit details
Commits on Jun 9, 2024
-
Configuration menu - View commit details
-
Copy full SHA for 696acaa - Browse repository at this point
Copy the full SHA 696acaaView commit details -
Configuration menu - View commit details
-
Copy full SHA for 0936406 - Browse repository at this point
Copy the full SHA 0936406View commit details
Commits on Jun 11, 2024
-
Configuration menu - View commit details
-
Copy full SHA for 636de1d - Browse repository at this point
Copy the full SHA 636de1dView commit details -
Configuration menu - View commit details
-
Copy full SHA for f52e234 - Browse repository at this point
Copy the full SHA f52e234View commit details
Commits on Jun 29, 2024
-
Add QTensor::quantize_onto (#12)
* Add the quantize_onto api * Take ref * Clippy * Format * Add error checking
Configuration menu - View commit details
-
Copy full SHA for bb8f6f0 - Browse repository at this point
Copy the full SHA bb8f6f0View commit details -
Configuration menu - View commit details
-
Copy full SHA for 5b04d96 - Browse repository at this point
Copy the full SHA 5b04d96View commit details -
Fix the fast bf16 gemm cublas kernels. (huggingface#2274)
* Use flash-attn in gemma. * Fix for the fast bf16 cublas gemm. * Fix some clippy lints. * Fix another lint. * Proper clippy fix.
Configuration menu - View commit details
-
Copy full SHA for f7095bb - Browse repository at this point
Copy the full SHA f7095bbView commit details -
Configuration menu - View commit details
-
Copy full SHA for b55b360 - Browse repository at this point
Copy the full SHA b55b360View commit details -
Depth Anything v2 (huggingface#2279)
* define structs * construct ResidualConvUnit * forward() for ResidualConvUnit * implement FeatureFusionBlock * implement Scratch * implement DPTHead * add identity module * implement forward for DTPHead * add get_intermediate_layers to DinoVisionTransformer * implement DepthAnythingV2 * some minor tweaks * fix compile errors * fix var builder prefixes * setup initial example * use fixed patch size of 37 (518 / 14) * debugged until output * print min and max values * add some dynamism to the output location * scale input image * extract prep function * extract output path function * normalize image with magic mean and std * add spectral coloring * squeeze in the right place * make enterpolation optional * use bail instead of panic * omit unnecessary Shape call * remove empty curly braces * use bail instead of assert * use vb and pp * remove closures * extract config object * Apply rustfmt. * Fix some clippy lints. * More lints. * Use the array methods. --------- Co-authored-by: laurent <[email protected]>
Configuration menu - View commit details
-
Copy full SHA for 08e93a6 - Browse repository at this point
Copy the full SHA 08e93a6View commit details -
Adding Gemm and ArgMax operators to candle-onnx (huggingface#2231)
* feat(gemm): implement Gemm operator in candle-onnx * feat(onnx): Add support for ArgMax operator in candle-onnx * Apply rustfmt. * Remove argmax as it was already present. --------- Co-authored-by: Laurent <[email protected]>
Configuration menu - View commit details
-
Copy full SHA for 5df1ae2 - Browse repository at this point
Copy the full SHA 5df1ae2View commit details -
Add DINOv2Reg4 + PlantCLEF2024 (huggingface#2293)
* Add: DINOv2Reg4 with PlantCLEF2024 weights and example ( See https://arxiv.org/abs/2309.16588 and https://zenodo.org/records/10848263 ) * Remove extra files + update README to download them + remove extra lines * minor fix (README remove extra spaces) * minor fix (README: Fix image url) * Modif: Add back interpolate_pos_encoding() + fix when no interpolation + remove extra comments + Update README ( source image changed and so the predictions ) * Fix: Improve code lisibility with '$ cargo clippy' and '$ cargo fmt' * Another clippy fix. --------- Co-authored-by: x-VEspit <[email protected]> Co-authored-by: laurent <[email protected]>
Configuration menu - View commit details
-
Copy full SHA for 0bb678c - Browse repository at this point
Copy the full SHA 0bb678cView commit details -
Configuration menu - View commit details
-
Copy full SHA for b438cba - Browse repository at this point
Copy the full SHA b438cbaView commit details
Commits on Jun 30, 2024
-
Configuration menu - View commit details
-
Copy full SHA for b7a3e34 - Browse repository at this point
Copy the full SHA b7a3e34View commit details
Commits on Jul 15, 2024
-
Configuration menu - View commit details
-
Copy full SHA for c967be9 - Browse repository at this point
Copy the full SHA c967be9View commit details
Commits on Jul 26, 2024
-
Configuration menu - View commit details
-
Copy full SHA for 9e09d7f - Browse repository at this point
Copy the full SHA 9e09d7fView commit details -
Configuration menu - View commit details
-
Copy full SHA for 8b357f6 - Browse repository at this point
Copy the full SHA 8b357f6View commit details
Commits on Jul 31, 2024
-
Configuration menu - View commit details
-
Copy full SHA for 2064fb0 - Browse repository at this point
Copy the full SHA 2064fb0View commit details
Commits on Aug 4, 2024
-
Configuration menu - View commit details
-
Copy full SHA for 1a48767 - Browse repository at this point
Copy the full SHA 1a48767View commit details -
Configuration menu - View commit details
-
Copy full SHA for 7bbcf00 - Browse repository at this point
Copy the full SHA 7bbcf00View commit details -
Configuration menu - View commit details
-
Copy full SHA for 1bf7101 - Browse repository at this point
Copy the full SHA 1bf7101View commit details -
Configuration menu - View commit details
-
Copy full SHA for d6d3d18 - Browse repository at this point
Copy the full SHA d6d3d18View commit details -
Configuration menu - View commit details
-
Copy full SHA for e20d85a - Browse repository at this point
Copy the full SHA e20d85aView commit details -
Configuration menu - View commit details
-
Copy full SHA for 8d2f32a - Browse repository at this point
Copy the full SHA 8d2f32aView commit details -
Configuration menu - View commit details
-
Copy full SHA for 9f144d6 - Browse repository at this point
Copy the full SHA 9f144d6View commit details -
Configuration menu - View commit details
-
Copy full SHA for c830f26 - Browse repository at this point
Copy the full SHA c830f26View commit details
Commits on Aug 5, 2024
-
Configuration menu - View commit details
-
Copy full SHA for 86d0876 - Browse repository at this point
Copy the full SHA 86d0876View commit details -
Configuration menu - View commit details
-
Copy full SHA for 8d8889c - Browse repository at this point
Copy the full SHA 8d8889cView commit details -
Configuration menu - View commit details
-
Copy full SHA for d18eb13 - Browse repository at this point
Copy the full SHA d18eb13View commit details -
Configuration menu - View commit details
-
Copy full SHA for d71b7d7 - Browse repository at this point
Copy the full SHA d71b7d7View commit details -
Configuration menu - View commit details
-
Copy full SHA for 412e9f4 - Browse repository at this point
Copy the full SHA 412e9f4View commit details
Commits on Aug 7, 2024
-
Configuration menu - View commit details
-
Copy full SHA for 27ca77e - Browse repository at this point
Copy the full SHA 27ca77eView commit details
Commits on Aug 9, 2024
-
* Add i32 dtype for cpu and cuda, with kernels * Fix cuda i32 * Fix cpu i32 * Add cuda map impls for i32 * Start to add to metal * Add the kernels * Oops * Fix dtype cast in safetensors * Oops * Oops * Add bf16 to i32 and vice versa casts
Configuration menu - View commit details
-
Copy full SHA for 7ad6494 - Browse repository at this point
Copy the full SHA 7ad6494View commit details
Commits on Aug 14, 2024
-
Configuration menu - View commit details
-
Copy full SHA for 6f0e190 - Browse repository at this point
Copy the full SHA 6f0e190View commit details -
Add the flux model for image generation. (huggingface#2390)
* Add the flux autoencoder. * Add the encoder down-blocks. * Upsampling in the decoder. * Sketch the flow matching model. * More flux model. * Add some of the positional embeddings. * Add the rope embeddings. * Add the sampling functions. * Add the flux example. * Fix the T5 bits. * Proper T5 tokenizer. * Clip encoder path fix. * Get the clip embeddings. * No configurable weights in layer norm. * More weights related fixes. * Yet another shape fix. * DType fix. * Fix a couple more shape issues. * DType fixes. * Fix the latent dims. * Fix more shape issues. * Autoencoder fixes. * Get some generations out. * Bugfix. * T5 padding. * Clippy fix. * Add the decode only mode. * Fix. * More fixes. * Finally get some generations to work. * Add readme.
Configuration menu - View commit details
-
Copy full SHA for ec55f58 - Browse repository at this point
Copy the full SHA ec55f58View commit details -
Configuration menu - View commit details
-
Copy full SHA for 0a146d7 - Browse repository at this point
Copy the full SHA 0a146d7View commit details -
Configuration menu - View commit details
-
Copy full SHA for 0f55c37 - Browse repository at this point
Copy the full SHA 0f55c37View commit details -
Configuration menu - View commit details
-
Copy full SHA for aef4eba - Browse repository at this point
Copy the full SHA aef4ebaView commit details -
Configuration menu - View commit details
-
Copy full SHA for c301efa - Browse repository at this point
Copy the full SHA c301efaView commit details -
add models support and example for THUDM/glm-4 (huggingface#2362)
* add models support and example for THUDM/glm-4 * fix the ci report * fmt * fix * Update README.org * Update README.org * fmt * Update README.org * README.md add codegeex4 * README.md add glm4 * Typo. * change expect into ? --------- Co-authored-by: Laurent Mazare <[email protected]>
Configuration menu - View commit details
-
Copy full SHA for fd0e933 - Browse repository at this point
Copy the full SHA fd0e933View commit details -
Add the MMDiT model of Stable Diffusion 3 (huggingface#2397)
* add mmdit of stable diffusion 3 lint add comments * correct a misplaced comment * fix cargo fmt * fix clippy error * use bail! instead of assert! * use get_on_dim in splitting qkv
Configuration menu - View commit details
-
Copy full SHA for f8e2b36 - Browse repository at this point
Copy the full SHA f8e2b36View commit details -
Configuration menu - View commit details
-
Copy full SHA for 0e78d29 - Browse repository at this point
Copy the full SHA 0e78d29View commit details -
fix: usage of
actions/checkout@v2
(huggingface#2403)* chore: changes from formatting on save * fix: usage of `actions/checkout@v2`
Configuration menu - View commit details
-
Copy full SHA for 1b796b9 - Browse repository at this point
Copy the full SHA 1b796b9View commit details -
Fix issues in the encodec example README.md (huggingface#2407)
Also squeeze the first dimension of the codes tensor in the example file to get the expected three dimensions.
Configuration menu - View commit details
-
Copy full SHA for c9cdd54 - Browse repository at this point
Copy the full SHA c9cdd54View commit details -
Soft Non-Maximum Suppression (huggingface#2400)
* Soft NMS with thresholds * NMS Test * Soft nms w/ boxes removed below threshold * Soft nms test * No longer removing bounding boxes to fit Soft-NMS focus * Initialize confidence * Added comments * Refactored out updating based on IOU/sigma * Score_threshold -> confidence_threshold for clarity * Remove bboxes below confidence threshold * Softnms basic functionality test * Softnms confidence decay test * Softnms confidence threshold test * Softnms no overlapping bbox test * Testing confidence after no overlap test * Single bbox and no bbox tests * Signify test completion * Handling result of test functions * Checking all pairs of bboxes instead of a forward pass * Equal confidence overlap test * Clarified tests for implementation * No longer dropping boxes, just setting to 0.0 * Formatted w/ cargo
Configuration menu - View commit details
-
Copy full SHA for 283a5cf - Browse repository at this point
Copy the full SHA 283a5cfView commit details -
Add documentation examples for
Tensor::i
andTensor::narrow
metho……ds (huggingface#2308) * Add documentation examples for `Tensor` methods * Apply fmt. * Cosmetic tweaks. --------- Co-authored-by: Laurent <[email protected]>
Configuration menu - View commit details
-
Copy full SHA for de719a2 - Browse repository at this point
Copy the full SHA de719a2View commit details -
Configuration menu - View commit details
-
Copy full SHA for 2e72a3d - Browse repository at this point
Copy the full SHA 2e72a3dView commit details -
Configuration menu - View commit details
-
Copy full SHA for d7a9bd0 - Browse repository at this point
Copy the full SHA d7a9bd0View commit details -
Clippy fixes. (huggingface#2415)
* Clippy fixes. * Bump the web_sys required version.
Configuration menu - View commit details
-
Copy full SHA for 3d40ffc - Browse repository at this point
Copy the full SHA 3d40ffcView commit details -
Configuration menu - View commit details
-
Copy full SHA for c5c5d49 - Browse repository at this point
Copy the full SHA c5c5d49View commit details -
Configuration menu - View commit details
-
Copy full SHA for 2386e4e - Browse repository at this point
Copy the full SHA 2386e4eView commit details -
Configuration menu - View commit details
-
Copy full SHA for a38053f - Browse repository at this point
Copy the full SHA a38053fView commit details
Commits on Aug 21, 2024
-
* Add GGUF bf16 type support * Add non avx impl for vec_dot_bf16 * Fix from_u32 * Fix loading * Fix dequant of bf16
Configuration menu - View commit details
-
Copy full SHA for 1b1974e - Browse repository at this point
Copy the full SHA 1b1974eView commit details
Commits on Aug 22, 2024
-
Configuration menu - View commit details
-
Copy full SHA for 36bd9f9 - Browse repository at this point
Copy the full SHA 36bd9f9View commit details -
Configuration menu - View commit details
-
Copy full SHA for 6fbddd6 - Browse repository at this point
Copy the full SHA 6fbddd6View commit details -
Add softcapping support to flash attention (#18)
* Expose the softcap methods * Add some tests * Fix generics
Configuration menu - View commit details
-
Copy full SHA for f706ef2 - Browse repository at this point
Copy the full SHA f706ef2View commit details
Commits on Sep 2, 2024
-
Update kernels for metal bf16 (#19)
* Update kernels for metal bf16 * Fix typo * Check if have bfloat
Configuration menu - View commit details
-
Copy full SHA for 3c8e120 - Browse repository at this point
Copy the full SHA 3c8e120View commit details
Commits on Sep 5, 2024
-
Configuration menu - View commit details
-
Copy full SHA for 014f140 - Browse repository at this point
Copy the full SHA 014f140View commit details
Commits on Sep 6, 2024
-
Configuration menu - View commit details
-
Copy full SHA for f317df8 - Browse repository at this point
Copy the full SHA f317df8View commit details -
onnx: workaround pow with negative base (huggingface#2439)
* onnx: workaround pow with negative base rather than fully defining pow in the cpu backend (as in huggingface#2318), this implements a much smaller change which is sufficient to evaluate silero-vad onnx models. Specifically, checking if pow is run with 2.0 exponent, and if so evaluate as simply `x*x` instead of the cpu backend of `e^(2.0 * ln(x))`. * PR: use Tensor::powf insead powf correctly handles a negative base.
Configuration menu - View commit details
-
Copy full SHA for 8a9d2be - Browse repository at this point
Copy the full SHA 8a9d2beView commit details -
onnx: support negative index in Gather (huggingface#2440)
index_select does not support negative indexing, but this change adds just enough workarounds in onnx to allow evaluating silero-vad models (which make use of negative indices).
Configuration menu - View commit details
-
Copy full SHA for a7142d3 - Browse repository at this point
Copy the full SHA a7142d3View commit details -
silero-vad v5 example (huggingface#2321)
* silero-vad v5 example This change adds an example of how to run silero-vad v5 * PR: rename 'vad' to 'silero-vad' * Update README.md --------- Co-authored-by: Laurent Mazare <[email protected]>
Configuration menu - View commit details
-
Copy full SHA for f62d7e8 - Browse repository at this point
Copy the full SHA f62d7e8View commit details -
Fix for parler-tts, do not add the last slice of padding tokens. (hug…
…gingface#2442) * Fix for parler-tts, do not add the last slice of padding tokens. * Support for the mini model.
Configuration menu - View commit details
-
Copy full SHA for ceab78e - Browse repository at this point
Copy the full SHA ceab78eView commit details -
Configuration menu - View commit details
-
Copy full SHA for 5b4c593 - Browse repository at this point
Copy the full SHA 5b4c593View commit details -
fix: qwen2 lm_head loading huggingface#2443 (huggingface#2445)
Co-authored-by: Yi Xu <[email protected]>
Configuration menu - View commit details
-
Copy full SHA for ef9649c - Browse repository at this point
Copy the full SHA ef9649cView commit details -
Update cudarc to 0.12. (huggingface#2451)
* Update cudarc to 0.12. * Some cudnn tweaks.
Configuration menu - View commit details
-
Copy full SHA for 7412bd0 - Browse repository at this point
Copy the full SHA 7412bd0View commit details -
FastViT fixes. (huggingface#2452)
* correct optional SE layer dimensions. * head_dim instead of num_heads is 32. * update test example output.
Configuration menu - View commit details
-
Copy full SHA for 8e39086 - Browse repository at this point
Copy the full SHA 8e39086View commit details -
MobileCLIP models S1 and S2 (huggingface#2454)
* Allow loading images with given std and mean * OpenCLIP text encoder component * Two MobileCLIP models * Clippy fixes. --------- Co-authored-by: Laurent <[email protected]>
Configuration menu - View commit details
-
Copy full SHA for 8632a2f - Browse repository at this point
Copy the full SHA 8632a2fView commit details -
Fix FLUX.1 weights (huggingface#2457)
* fix FLUX.1 weights * added flux1-dev.safetensors
Configuration menu - View commit details
-
Copy full SHA for f492c04 - Browse repository at this point
Copy the full SHA f492c04View commit details -
Clippy fixes for 1.81.0. (huggingface#2461)
* Clippy fixes for 1.81.0. * Another fix.
Configuration menu - View commit details
-
Copy full SHA for 91e0c6e - Browse repository at this point
Copy the full SHA 91e0c6eView commit details
Commits on Sep 11, 2024
-
Improve candle_core::Error to make it more ergonomic (#21)
* Bump the version to 0.6.1. (huggingface#2438) * onnx: workaround pow with negative base (huggingface#2439) * onnx: workaround pow with negative base rather than fully defining pow in the cpu backend (as in huggingface#2318), this implements a much smaller change which is sufficient to evaluate silero-vad onnx models. Specifically, checking if pow is run with 2.0 exponent, and if so evaluate as simply `x*x` instead of the cpu backend of `e^(2.0 * ln(x))`. * PR: use Tensor::powf insead powf correctly handles a negative base. * onnx: support negative index in Gather (huggingface#2440) index_select does not support negative indexing, but this change adds just enough workarounds in onnx to allow evaluating silero-vad models (which make use of negative indices). * silero-vad v5 example (huggingface#2321) * silero-vad v5 example This change adds an example of how to run silero-vad v5 * PR: rename 'vad' to 'silero-vad' * Update README.md --------- Co-authored-by: Laurent Mazare <[email protected]> * Fix for parler-tts, do not add the last slice of padding tokens. (huggingface#2442) * Fix for parler-tts, do not add the last slice of padding tokens. * Support for the mini model. * Add FastViT model. (huggingface#2444) * fix: qwen2 lm_head loading huggingface#2443 (huggingface#2445) Co-authored-by: Yi Xu <[email protected]> * Update cudarc to 0.12. (huggingface#2451) * Update cudarc to 0.12. * Some cudnn tweaks. * FastViT fixes. (huggingface#2452) * correct optional SE layer dimensions. * head_dim instead of num_heads is 32. * update test example output. * MobileCLIP models S1 and S2 (huggingface#2454) * Allow loading images with given std and mean * OpenCLIP text encoder component * Two MobileCLIP models * Clippy fixes. --------- Co-authored-by: Laurent <[email protected]> * Fix FLUX.1 weights (huggingface#2457) * fix FLUX.1 weights * added flux1-dev.safetensors * Clippy fixes for 1.81.0. (huggingface#2461) * Clippy fixes for 1.81.0. * Another fix. * Make Error::msg more in line with anyhow::Error::msg * Add context trait * Even more flexible * Format --------- Co-authored-by: Laurent Mazare <[email protected]> Co-authored-by: shua <[email protected]> Co-authored-by: Jani Monoses <[email protected]> Co-authored-by: ilookee <[email protected]> Co-authored-by: Yi Xu <[email protected]> Co-authored-by: Eugene Hauptmann <[email protected]>
Configuration menu - View commit details
-
Copy full SHA for ad84486 - Browse repository at this point
Copy the full SHA ad84486View commit details -
Add API to get current device seed (#22)
* Add api to get current seed * Remove cell for rwlock
Configuration menu - View commit details
-
Copy full SHA for 7f5a470 - Browse repository at this point
Copy the full SHA 7f5a470View commit details
Commits on Sep 13, 2024
-
Configuration menu - View commit details
-
Copy full SHA for 9240d03 - Browse repository at this point
Copy the full SHA 9240d03View commit details -
Configuration menu - View commit details
-
Copy full SHA for 8a99f7c - Browse repository at this point
Copy the full SHA 8a99f7cView commit details
Commits on Sep 15, 2024
-
Configuration menu - View commit details
-
Copy full SHA for 9e31a19 - Browse repository at this point
Copy the full SHA 9e31a19View commit details
Commits on Oct 2, 2024
-
Configuration menu - View commit details
-
Copy full SHA for d08212c - Browse repository at this point
Copy the full SHA d08212cView commit details -
Configuration menu - View commit details
-
Copy full SHA for c04861d - Browse repository at this point
Copy the full SHA c04861dView commit details -
Configuration menu - View commit details
-
Copy full SHA for 156ebd1 - Browse repository at this point
Copy the full SHA 156ebd1View commit details
Commits on Oct 3, 2024
-
Configuration menu - View commit details
-
Copy full SHA for 20a57c4 - Browse repository at this point
Copy the full SHA 20a57c4View commit details
Commits on Oct 6, 2024
-
Configuration menu - View commit details
-
Copy full SHA for 121bdfd - Browse repository at this point
Copy the full SHA 121bdfdView commit details