Commits

Breadcrumbs

History for

onmain

User selector

All users

Datepicker

All time

Commit History

Commits on Nov 22, 2024

• Reinstate llama tests (#141)

  alessandropalla

  authored
  c779203

  Copy full SHA for c779203

Commits on Nov 14, 2024

• Building on Ubuntu 24.04 (#129)

  Show description for 0a449c5
  ytxmobile98

  authored
  0a449c5

  Copy full SHA for 0a449c5

Commits on Jul 25, 2024

• Add prelu and normalize ops (#107)

  alessandropalla

  authored
  f04d499

  Copy full SHA for f04d499

• Remove assert check for pythion 3.12

  alessandropalla

  committed
  f3b112a

  Copy full SHA for f3b112a

Commits on Jul 23, 2024

• Dtype fix for backprop matmul (#104)

  Show description for 86e4e1e
  

  SarahByrneIntel

  and

  SarahByrneIntel

  authored
  86e4e1e

  Copy full SHA for 86e4e1e

Commits on Jul 20, 2024

• Fix config.use_to for npu compile (#100)

  xduzhangjiayu

  authored
  0562410

  Copy full SHA for 0562410

Commits on Jul 19, 2024

• Support for Phi-3 MLP layer (#84)

  Show description for 2193535
  

  3 peopleauthored
  2193535

  Copy full SHA for 2193535

Commits on Jul 5, 2024

• Make matmul op (@) torch compliant (#91)

  Show description for 6f8e11d
  alessandropalla

  authored
  6f8e11d

  Copy full SHA for 6f8e11d

• Adding support and testing for chunk tensor operation (#90)

  Show description for 13b9e3d
  

  SarahByrneIntel

  and

  SarahByrneIntel

  authored
  13b9e3d

  Copy full SHA for 13b9e3d

Commits on Jul 4, 2024

• Fix ops and r_ops in case of float and int (#88)

  Show description for 66c1205
  alessandropalla

  authored
  66c1205

  Copy full SHA for 66c1205

Commits on Jul 1, 2024

• Adding support for operations on tensors (#81)

  Show description for 0ac709b
  

  SarahByrneIntel

  and

  SarahByrneIntel

  authored
  0ac709b

  Copy full SHA for 0ac709b

• Adding power and log softmax operations (#80)

  Show description for 641f730
  

  SarahByrneIntel

  and

  SarahByrneIntel

  authored
  641f730

  Copy full SHA for 641f730

Commits on Jun 26, 2024

• version v1.3.0

  alessandropalla

  committed
  e1d9155

  Copy full SHA for e1d9155

• Fix SDPA in case attn_mask == None (#78)

  alessandropalla

  authored
  75bc3c9

  Copy full SHA for 75bc3c9

• Hotfix for module (#76)

  Show description for 736afb5
  alessandropalla

  authored
  736afb5

  Copy full SHA for 736afb5

Commits on Jun 25, 2024

• Sarah/feature/reduce ops (#74)

  Show description for 4f17015
  

  SarahByrneIntel

  and

  SarahByrneIntel

  authored
  4f17015

  Copy full SHA for 4f17015

• Automatic handling of output layers (#73)

  Show description for 56dd4d7
  alessandropalla

  authored
  56dd4d7

  Copy full SHA for 56dd4d7

Commits on Jun 24, 2024

• Add torch NPU device (#72)

  Show description for 1f2410e
  alessandropalla

  authored
  1f2410e

  Copy full SHA for 1f2410e

• Sarah/feature/operations (#68)

  Show description for 605aaa5
  

  3 peopleauthored
  605aaa5

  Copy full SHA for 605aaa5

• fix BatchNorm layer (#71)

  Show description for 0569ada
  alessandropalla

  authored
  0569ada

  Copy full SHA for 0569ada

• Add torch.nn.functional.conv2d (#70)

  Show description for ed0993b
  alessandropalla

  authored
  ed0993b

  Copy full SHA for ed0993b

• Better torch integration (#66)

  Show description for b67bd8b
  alessandropalla

  authored
  b67bd8b

  Copy full SHA for b67bd8b

Commits on Jun 21, 2024

• Adding support for L2 normalisation operation (#65)

  Show description for e4d0d61
  

  SarahByrneIntel

  and

  SarahByrneIntel

  authored
  e4d0d61

  Copy full SHA for e4d0d61

Commits on Jun 20, 2024

• Add memory operation and tensor class (#63)

  Show description for 417bd76
  alessandropalla

  authored
  417bd76

  Copy full SHA for 417bd76

• Sarah/feature/constant operation support (#62)

  Show description for 8e2172d
  

  SarahByrneIntel

  and

  SarahByrneIntel

  authored
  8e2172d

  Copy full SHA for 8e2172d

Commits on Jun 19, 2024

• Native convolution and dw convolution (#61)

  Show description for fa4c82a
  alessandropalla

  authored
  fa4c82a

  Copy full SHA for fa4c82a

Commits on Jun 18, 2024

• Change function generate_with_static_shape (#60)

  xduzhangjiayu

  authored
  bcb1315

  Copy full SHA for bcb1315

• Fix for NaNs in LLM inference (#58)

  Show description for cf896af
  alessandropalla

  authored
  cf896af

  Copy full SHA for cf896af

• adding new operations (#50)

  Show description for 1e9a533
  

  3 peopleauthored
  1e9a533

  Copy full SHA for 1e9a533

Commits on Jun 11, 2024

• Fix int4 quantization for llama and gemma (#47)

  Show description for c486f74
  alessandropalla

  authored
  c486f74

  Copy full SHA for c486f74

Commits on Jun 6, 2024

• Create warnings if library is not present (#46)

  alessandropalla

  authored
  141d346

  Copy full SHA for 141d346

• Better driver versioning system

  alessandropalla

  committed
  5de9e69

  Copy full SHA for 5de9e69

Commits on Jun 5, 2024

• Fix export error with trust_remote_code (#43)

  Show description for 61e0d16
  alessandropalla

  authored
  61e0d16

  Copy full SHA for 61e0d16

Commits on Jun 4, 2024

• Persistent compilation (#39)

  Show description for c26443e
  alessandropalla

  authored
  c26443e

  Copy full SHA for c26443e

• Implement SDPA (Scalar dot product attention) NPU kernel (#36)

  Show description for d4928c4
  alessandropalla

  authored
  d4928c4

  Copy full SHA for d4928c4

Pagination

Previous

Next