summaryrefslogtreecommitdiff
path: root/candle-examples
Commit message (Collapse)AuthorAgeFilesLines
...
* Fix FLUX.1 weights (#2457)Eugene Hauptmann2024-08-291-3/+3
| | | | | * fix FLUX.1 weights * added flux1-dev.safetensors
* MobileCLIP models S1 and S2 (#2454)Jani Monoses2024-08-294-10/+250
| | | | | | | | | | | | | * Allow loading images with given std and mean * OpenCLIP text encoder component * Two MobileCLIP models * Clippy fixes. --------- Co-authored-by: Laurent <laurent.mazare@gmail.com>
* FastViT fixes. (#2452)Jani Monoses2024-08-281-5/+5
| | | | | * correct optional SE layer dimensions. * head_dim instead of num_heads is 32. * update test example output.
* Add FastViT model. (#2444)Jani Monoses2024-08-232-0/+122
|
* Fix for parler-tts, do not add the last slice of padding tokens. (#2442)Laurent Mazare2024-08-221-2/+21
| | | | | * Fix for parler-tts, do not add the last slice of padding tokens. * Support for the mini model.
* silero-vad v5 example (#2321)shua2024-08-223-0/+216
| | | | | | | | | | | | | * 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 <laurent.mazare@gmail.com>
* Update README.md (#2435)Laurent Mazare2024-08-191-1/+3
|
* Add a readme for the parler-tts example. (#2434)Laurent Mazare2024-08-193-30/+21
| | | | | | | | | * Add a readme for the parler-tts example. * Remove the python decode script. * mp4 tweaks. * Another readme tweak.
* Add the DAC model. (#2433)Laurent Mazare2024-08-192-7/+20
| | | | | | | | | * Add the DAC model. * More quantization support. * Handle DAC decoding. * Plug the DAC decoding in parler-tts.
* parler-tts support (#2431)Laurent Mazare2024-08-182-0/+204
| | | | | | | | | | | | | | | | | | | | | | | * Start sketching parler-tts support. * Implement the attention. * Add the example code. * Fix the example. * Add the description + t5 encode it. * More of the parler forward pass. * Fix the positional embeddings. * Support random sampling in generation. * Handle EOS. * Add the python decoder. * Proper causality mask.
* Fix the marian tokenizer importer. (#2426)Laurent Mazare2024-08-171-4/+16
| | | | | * Fix the marian tokenizer importer. * Ignore the python caches.
* Add support for gemma-2. (#2425)Laurent Mazare2024-08-172-22/+74
| | | | | | | | | | | * Add gemma-2. * Support a couple more models. * Sliding window support. * Example + readme updates. * Update the main readme.
* Apply rustfmt. (#2421)Laurent Mazare2024-08-161-1/+0
|
* Fix build issue in EOS Token in llama-multiprocess (#2420)Hadi2024-08-161-2/+11
|
* Add Based LLM from Hazy Research. (#2411)Jani Monoses2024-08-122-0/+295
|
* Fix issues in the encodec example README.md (#2407)Joel Nises2024-08-102-1/+1
| | | Also squeeze the first dimension of the codes tensor in the example file to get the expected three dimensions.
* Add the import script for the T5 tokenizer. (#2399)Laurent Mazare2024-08-051-0/+6
|
* add models support and example for THUDM/glm-4 (#2362)唐璜2024-08-052-0/+332
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | * 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 <laurent.mazare@gmail.com>
* Support for mistral-nemo. (#2396)Laurent Mazare2024-08-041-7/+14
|
* Support the flux-dev model too. (#2395)Laurent Mazare2024-08-041-9/+37
|
* Add the flux model for image generation. (#2390)Laurent Mazare2024-08-043-0/+201
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * 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.
* Fix cargo fmt. (#2383)Laurent Mazare2024-08-011-14/+19
| | | | | | | * Fix cargo fmt. * Clippy fix. * Cosmetic tweaks.
* Jina Bert Example fix and more configuration (#2191)Joan Fontanals2024-08-011-12/+28
| | | | | | | * fix: fix jina bert example logic * feat: enable jina embeddings de * feat: allow more flexibility on Jina Bert
* Add Hiera vision model. (#2382)Jani Monoses2024-08-012-0/+117
|
* Enable BF16 on metal. (#2380)Laurent Mazare2024-08-011-4/+2
|
* Use BF16 on metal when possible. (#2378)Laurent Mazare2024-08-011-5/+1
|
* bert attention mask (#1934)Zheng Li2024-08-011-2/+10
| | | | | | | | | | | | | | | * bert attention mask * Allow for using None as a mask. * Revert part of the changes so that the proper default mask applies. * Cosmetic change. * Another cosmetic tweak. --------- Co-authored-by: Laurent <laurent.mazare@gmail.com>
* Add support for Llama 3.1 (#2359)Eric Buehler2024-07-263-8/+26
| | | | | | | | | | | | | | | | | * Add Llama 3.1 rope * Clippy * Format * Clippy * Add support for multiple eos tokens: * Untagged either * Remove either dep and fix settings.json * Make the max positional embeddings configurable
* onnx: fix pad, unsqueeze (#2317)shua2024-07-2313-14/+14
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * onnx: fix pad, unsqueeze both implementations have off-by-one errors: - Pad 'reflect' cycle for eg `dim==3` is `[0,1,2,1]` which has length of 4 (or `dim*2 - 2`) not 5 (current code `dim*2 - 1`) - Unsqueeze(-1) for tensor with `dim==3` should be 3 (ie `dim+index+1`) not 2 (ie currently `dim+index`) in addition, Pad is incorrectly calculating the starting padding. If we want to pad out 2 elements to the start, and we have this cycle of indices of length 6, then we should skip 4 elements, but currently we skip 2. A more visual representation of what's going on is below: ``` pad_start: 2 data: [a,b,c,d] indices: [0, 1, 2, 3, 2, 1, 0, 1, 2, 3, 2, 1, 0, ..] // zigzag between 0..4 actual: skip [ c d| c b a b] expected: ~ skip ~ [ c b| a b c d] ``` The values between `[` and `|` are padding and the values between `|` and `]` in the example should match the original data being padded. * Fix clippy lints. --------- Co-authored-by: Laurent <laurent.mazare@gmail.com>
* fix clip example title (#2345)Caio Petrucci Rosa2024-07-231-1/+1
|
* feat(candle-transformers/models/codegeex4-9b): add codegeex4-9 (#2334)donjuanplatinum2024-07-212-0/+348
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * feat(candle-transformers/models/codegeex4-9b): add codegeex4-9b transoformers * change mod.rs * feat(candle-examples/codegeex4-9b) * Update codegeex4_9b.rs * Update main.rs * Update codegeex4_9b.rs * Update main.rs * fmt * fix * fmt * Clippy fix. * Remove some print statements. * Avoid using unwrap. * 1. add README 2. change the print fmt * Another clippy fix. --------- Co-authored-by: Laurent <laurent.mazare@gmail.com>
* Pin the revision used by moondream. (#2340)Laurent Mazare2024-07-181-7/+15
|
* Add mathstral in the examples. (#2339)Laurent Mazare2024-07-181-0/+3
|
* add quantized qwen2 (#2329)Zhuo Jinggang2024-07-122-0/+317
| | | | | * add quantized version of qwen2 and corresponding example for qwen2-instruct * fix quantized qwen2 clippy error
* Add Mobilenet v4 (#2325)Jani Monoses2024-07-093-17/+137
| | | | | | | * Support different resolutions in load_image() * Added MobilenetV4 model. * Add MobileNetv4 to README
* Add EVA-02 model ( https://arxiv.org/abs/2303.11331 ) (#2311)v-espitalier2024-07-072-0/+103
| | | | | | | | | | | | * Add EVA-02 model ( https://arxiv.org/abs/2303.11331 ) * Clippy fix. * And apply fmt. --------- Co-authored-by: v-espitalier <> Co-authored-by: Laurent <laurent.mazare@gmail.com>
* Beit: Add the gen_relative_position_index() function (#2306)v-espitalier2024-07-041-1/+1
| | | Co-authored-by: v-espitalier <>
* Add Beit model ( https://arxiv.org/abs/2106.08254 ) (#2305)v-espitalier2024-07-012-0/+99
| | | Co-authored-by: v-espitalier <>
* make up for the missing last token output of phi2 example (#2299)Czxck0012024-06-291-0/+4
|
* Add DINOv2Reg4 + PlantCLEF2024 (#2293)v-espitalier2024-06-293-0/+113
| | | | | | | | | | | | | | | | | | | | * 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 <vincent.espitalier@cirad.fr> Co-authored-by: laurent <laurent.mazare@gmail.com>
* Depth Anything v2 (#2279)Jeroen Vlek2024-06-244-0/+257
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * 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 <laurent.mazare@gmail.com>
* Support for the new Qwen2 models. (#2257)Laurent Mazare2024-06-071-10/+26
| | | | | * Support for the new Qwen2 models. * Add more models.
* Add LLaVA support (#2234)chenwanqq2024-06-035-0/+791
| | | | | | | | | | | | | | | | | | | | | | | | | * first commit * llava * clippy and fmt * some fixes * minor fixes * remove useless file * refactor: Remove llava/constants.rs and update llava/mod.rs * modify variable name * modify code after clippy * Minor tweaks. --------- Co-authored-by: laurent <laurent.mazare@gmail.com>
* Simplify the KvCache api. (#2207)Laurent Mazare2024-05-231-1/+0
|
* Add Phi-3 Medium (#2205)Jani Monoses2024-05-231-6/+13
|
* Use flash-attn in gemma. (#2195)Laurent Mazare2024-05-181-1/+4
| | | | | * Use flash-attn in gemma. * Fix flash-attn for head dim 256.
* Support flash-attn in quantized phi3. (#2194)Laurent Mazare2024-05-181-1/+10
|
* Add a slice_set op. (#2193)Laurent Mazare2024-05-181-1/+1
| | | | | | | | | | | | | | | * Add a slice_set op. * Add some testing. * Add the dedicated kv-cache module. * Derive debug and clone. * Expose more kv-cache functions. * Return the current data when appending. * Use the new cache in the quantized phi3 model.
* Support embedding model gte-Qwen1.5-7B-instruct (#2190)Yin Guobing2024-05-163-1/+198
| | | | | | | | | | | | | | | | | * Support embedding model gte-Qwen1.5-7B-instruct This is a text embedding model based on Qwen2. They share same model architecture except the last MLP module. This commit brings in minimal modification of the old Qwen2 implementation to support both models. An example is provided, and had been verified according to the official PyTorch implementation. * Avoid doing the 'last-token filtering' based on the absence of attention mask. --------- Co-authored-by: Laurent <laurent.mazare@gmail.com>
* Allow the threshold argumet to be negative in the segment-anything example ↵Daniel Varga2024-05-151-1/+1
| | | | | | | | | | (#2187) Threshold is 0.0 by default, negative values make more points included, expanding the mask. Positive values make it more picky, making the mask smaller. Negative numbers start with a minus sign, which normally makes clap consider it a flag.
generated by cgit v1.2.3 (git 2.39.1) at 2025-04-05 01:39:01 +0000