Commit Graph

  • cc06ba2294 fix bad pattern matching and function name Baye Dieng 2023-12-29 09:46:24 +00:00
  • a6bd0b47a5 Fix the CI. Nicolas Patry 2023-12-29 10:17:52 +01:00
  • b59b1b2bb6 remove generated png Baye Dieng 2023-12-28 21:50:58 +00:00
  • 3922b42c18 add urecip op to metal backend Baye Dieng 2023-12-28 21:50:12 +00:00
  • 1e442d4bb9 Fix lints for clippy 1.75. (#1494) Laurent Mazare 2023-12-28 20:26:20 +01:00
  • cd889c0f8a add config_amazon_mistral_lite (#1493) Daniel Clough 2023-12-28 10:59:58 -08:00
  • 8e93e76a91 fixes error message Gonzalo 2023-12-28 15:03:05 -03:00
  • b3e838f3e2 cargo fmt Gonzalo 2023-12-28 14:07:34 -03:00
  • 289c57d600 Removing metal fences. Increases performance substantially on m1 pro. ivarflakstad/metal-fenceless Ivar Flakstad 2023-12-28 17:31:07 +01:00
  • 8bf892403a Improves metal's not implemented error messages Gonzalo 2023-12-28 11:04:06 -03:00
  • d35f0a1376 Bump the crate version to 0.3.3. (#1490) Laurent Mazare 2023-12-28 13:38:30 +01:00
  • 65cb90bd40 Add some mention to SOLAR-10.7B in the readme. (#1487) Laurent Mazare 2023-12-27 15:25:39 +01:00
  • 996a7f2e24 Rework the llama example config, add the solar model. (#1485) Laurent Mazare 2023-12-26 22:24:04 +01:00
  • 3071ea6c3e Use the new hub helper function. (#1484) Laurent Mazare 2023-12-26 09:44:30 +01:00
  • 37c539f2b7 Helper function to load sharded safetensors files (#1481) Laurent Mazare 2023-12-25 21:49:21 +01:00
  • eae3a20d43 Merge pull request #1479 from huggingface/upsample_metal Nicolas Patry 2023-12-25 14:25:53 +01:00
  • 13a5d15ebc Adding upsample_nearest_2d. Nicolas Patry 2023-12-25 14:25:19 +01:00
  • 1505d85276 Merge pull request #1461 from huggingface/metal-conv Nicolas Patry 2023-12-25 12:48:09 +01:00
  • 95e18ef675 Fixing matmul for convolutions. Nicolas Patry 2023-12-25 12:29:34 +01:00
  • 7135791dd5 Fix the quantized mistral example. (#1478) Laurent Mazare 2023-12-25 09:31:24 +01:00
  • 88589d8815 Support mistral instruct v0.2. (#1475) Laurent Mazare 2023-12-23 16:18:49 +01:00
  • 5b35fd0fcf MMLU evaluation for Phi. (#1474) Laurent Mazare 2023-12-23 15:28:36 +01:00
  • ba1fae590e Validate the kernel size in pooling ops. (#1473) Laurent Mazare 2023-12-23 11:19:22 +01:00
  • 78d982e1bd Fix for mamba 2.8b. (#1472) Laurent Mazare 2023-12-23 11:01:39 +01:00
  • d8b9a727fc Support different mamba models. (#1471) Laurent Mazare 2023-12-23 10:46:02 +01:00
  • ceb78d3e28 Sketch the minimal mamba example. (#1465) Laurent Mazare 2023-12-22 00:28:50 +01:00
  • f6408a3779 feat: add clear_kv_cache to mistral and qmistral models (#1464) drbh 2023-12-21 15:19:19 -05:00
  • 10d94659c3 Adding the convolutions (1d + 2d) to candle on metal. Nicolas Patry 2023-12-21 10:39:24 +01:00
  • 563a79afa1 make fn name generic (#1459) Daniel Clough 2023-12-20 17:16:31 -08:00
  • 8ede5f4210 add fn config_chat_ml (#1458) Daniel Clough 2023-12-20 12:03:24 -08:00
  • 9fc210fae8 Merge pull request #1318 from huggingface/metal4 Nicolas Patry 2023-12-20 15:37:31 +01:00
  • 9b5e4843a6 Optimizing decode matmul (Phi at 28tok/s on M3). Nicolas Patry 2023-12-20 09:54:19 +01:00
  • 5edb07a5b1 mps matmul metal4.7-mps Ivar Flakstad 2023-12-20 02:53:18 +01:00
  • 03641293ee Clippy pass. metal4-m3 Nicolas Patry 2023-12-18 15:22:43 +01:00
  • 064ba17bd7 Remove print. Nicolas Patry 2023-12-18 11:04:16 +01:00
  • e8ee253ee0 Missing cast. Nicolas Patry 2023-12-18 11:01:18 +01:00
  • 8bd3d6b94b Index add. Nicolas Patry 2023-12-18 10:46:01 +01:00
  • 6a3ca7da0c Scatter add. Nicolas Patry 2023-12-18 10:32:22 +01:00
  • 96f1a28e39 Add a simple full method. (#1455) Laurent Mazare 2023-12-17 20:15:57 -05:00
  • 586b6f6fff Adding gather op. Nicolas Patry 2023-12-17 23:34:12 +01:00
  • e4b0cc59f5 Adding CMP Nicolas Patry 2023-12-17 22:32:25 +01:00
  • 0a6e0a8c9a Implement randn (CPU-> device) Nicolas Patry 2023-12-17 19:09:08 +01:00
  • 972903021c Finish reduce kernels. Nicolas Patry 2023-12-17 19:07:00 +01:00
  • 94817dac56 Bump the crate version to 0.3.2. (#1452) Laurent Mazare 2023-12-17 05:34:53 -06:00
  • 1e86717bf2 Fix a couple typos (#1451) Laurent Mazare 2023-12-17 05:20:05 -06:00
  • c630622a07 Expose AdamW parameters (#1449) Dave Lage 2023-12-16 19:41:56 -05:00
  • c4cfcf1539 Tweak the readme for phi and the default sample length. (#1450) Laurent Mazare 2023-12-16 18:11:36 -06:00
  • 1782e93de6 Mixtral quantized instruct. (#1447) Laurent Mazare 2023-12-16 16:16:39 -06:00
  • cfdf9640a3 Readme tweaks. (#1446) Laurent Mazare 2023-12-16 06:23:12 -06:00
  • e12cbfd73b Update the readme to mention mixtral. (#1443) Laurent Mazare 2023-12-15 19:29:03 -06:00
  • 30a958e5dd Quantized mixtral model (#1442) Laurent Mazare 2023-12-15 19:16:06 -06:00
  • 614842b311 Add the Mixtral model. (#1437) Laurent Mazare 2023-12-15 14:19:56 -06:00
  • 79eab519fd Fix phi example (#1436) niu tech 2023-12-15 14:01:10 +01:00
  • 6bc92e63cb Addressing a lot of comments. Nicolas Patry 2023-12-15 13:06:04 +01:00
  • aa04015098 Remove unwrap(). Nicolas Patry 2023-12-15 12:23:28 +01:00
  • 8b5059e951 Remove test file. Nicolas Patry 2023-12-15 11:55:30 +01:00
  • 26540641c1 Renamed all kernel names. Nicolas Patry 2023-12-15 11:24:47 +01:00
  • 34d83377f6 Better error message on older macos Nicolas Patry 2023-12-15 11:18:54 +01:00
  • 77197379cc More cleanup. Nicolas Patry 2023-12-15 11:17:05 +01:00
  • 916a8c5464 Revert candle-transformers. Nicolas Patry 2023-12-15 11:15:21 +01:00
  • 243e83f2b9 Adding a bunch of docs ! Nicolas Patry 2023-12-15 11:02:41 +01:00
  • cf27868b57 More cleanup. metal4.7 Nicolas Patry 2023-12-15 01:44:22 +01:00
  • 40c3e1bd5a cleanup. Nicolas Patry 2023-12-15 01:41:14 +01:00
  • ece4c69a68 Fixing softmax. Nicolas Patry 2023-12-15 01:35:08 +01:00
  • 4eeaf205d6 Fix softmax for long sequences (missing barrier). Nicolas Patry 2023-12-14 19:37:03 +01:00
  • f419a38e1a Fix use resource. Nicolas Patry 2023-12-14 16:52:37 +01:00
  • 361f2ad2af Working with merging encoders and using fences. Nicolas Patry 2023-12-14 16:05:33 +01:00
  • e60f9b5dfc Speedup ShardedSafeTensors to load Tensors with default hints (#1384) YiiSh 2023-12-14 22:08:56 +08:00
  • 7be982f6f7 Mention phi-2 in the readme. (#1434) Laurent Mazare 2023-12-14 08:02:27 -06:00
  • 104e196d46 Phi 2 wasm (#1432) Radamés Ajna 2023-12-14 04:04:17 -08:00
  • 5e33c85c8f Quantized version for phi-v2. (#1430) Laurent Mazare 2023-12-13 21:16:34 -06:00
  • 2b3a018be7 Support for phi-2. (#1429) Laurent Mazare 2023-12-13 20:59:29 -06:00
  • 931432ed55 Fixing tests + matmul from MFA Nicolas Patry 2023-12-13 16:58:36 +01:00
  • 0404a3eb5b Removed MPSMatrix entirely (buggy). Nicolas Patry 2023-12-13 16:21:48 +01:00
  • 1f23cea90c MFA metal4.5 Nicolas Patry 2023-12-13 16:09:20 +01:00
  • a9d0657432 Better version ? metal4.6 Nicolas Patry 2023-12-13 12:09:20 +01:00
  • 4cb443d00a Fix the logsumexp test. (#1426) Laurent Mazare 2023-12-12 10:56:11 -06:00
  • 87dc559817 Lots of updates including some stack of command buffers. nicolas 2023-12-12 17:41:56 +01:00
  • 77252ffb82 Add logsumexp function (#1424) Wenqing Zong 2023-12-12 16:32:17 +00:00
  • a0282751d5 Tmp. tmp4 nicolas 2023-12-11 19:51:46 +01:00
  • ce33d6ad2a Tmp. nicolas 2023-12-11 11:10:48 +01:00
  • 3d0ade406a Tmp. nicolas 2023-12-11 09:38:25 +01:00
  • ce0783d9ff Stash for debugging metal4-mfa Ivar Flakstad 2023-12-10 13:11:53 +01:00
  • 18eb87f25f Upsample grad (#1420) KGrewal1 2023-12-10 07:43:24 +00:00
  • da0af3cb3e Merge pull request #1408 from jbochi/metal_gelu2 Nicolas Patry 2023-12-09 19:46:36 +01:00
  • 35352e441a Begin adding mfa support Ivar Flakstad 2023-12-08 21:51:49 +01:00
  • 9bd94c1ffa Speed up bert with approx gelu (#1410) Juarez Bochi 2023-12-06 11:46:37 -05:00
  • 803ac8405b Put back affine strided tests Nicolas Patry 2023-11-30 11:40:39 +01:00
  • 6e25822d4f Fix gelu for large x Juarez Bochi 2023-12-06 09:59:44 -05:00
  • 236b820e28 Another prelu bugfix. (#1407) Laurent Mazare 2023-12-06 09:54:41 +01:00
  • 2648e797c2 Use the proper broadcasting for prelu. (#1406) Laurent Mazare 2023-12-05 07:09:31 +01:00
  • b5c283e86f Add the prelu layer. (#1402) Laurent Mazare 2023-12-03 17:06:09 +01:00
  • 8418154ee0 Add nvcc ccbin support to examples (#1401) emka 2023-12-03 17:01:16 +01:00
  • 99b7273b03 Add compute cap env support to examples (#1400) emka 2023-12-03 17:00:24 +01:00
  • 16161145ae Add the leo models to the quantized examples. (#1398) Laurent Mazare 2023-12-03 13:30:41 +01:00
  • 0738df5290 Add more mentions to SDXL Turbo in the readme. (#1397) Laurent Mazare 2023-12-03 11:41:21 +01:00
  • 37bf1ed012 Stable Diffusion Turbo Support (#1395) Edwin Cheng 2023-12-03 15:37:10 +08:00
  • dd40edfe73 Add Euler Ancestral Discrete Scheduler (#1390) Edwin Cheng 2023-12-03 03:59:23 +08:00
  • 5aa1a65dab Add quantized Starling, fix open-chat prompt (#1393) Lucas de Ávila Martins 2023-12-02 13:47:19 -03:00
  • 2ca086939f Put back affine strided tests Nicolas Patry 2023-11-30 11:40:39 +01:00