Commits

Breadcrumbs

History for

on3974e55

User selector

All users

Datepicker

All time

Commit History

Commits on Jul 16, 2023

• fix(llm): pass ModelParameters to InferenceSession

  philpax

  committed
  06ad53e

• fix(llm): pass use_gpu to InferenceSession

  philpax

  committed
  fd10a35

• fix rustformers#369 - clean up tensor name interface

  philpax

  committed
  dfede25

• Merge pull request rustformers#325 from LLukas22/feat/cuda-opencl-acceleration

  Show description for 3062a08
  philpax

  authored
  3062a08

• feat(tracing): add tracing and update CLI to use tracing

  Show description for c344592
  radu-matei

  committed
  c344592

Commits on Jul 15, 2023

• refactor(ggml): accelerator/tensor/context

  philpax

  committed
  e7ac55b

• refactor: remove unused tensorloader tensors

  philpax

  committed
  67bbaf9

• Review fixes

  LLukas22

  committed
  cea02f2

Commits on Jul 12, 2023

• Update llama.cpp and integrate graph "planning"

  LLukas22

  committed
  53095b1

• Merge remote-tracking branch 'upstream/main' into feat/cuda-opencl-acceleration

  LLukas22

  committed
  7c0a29a

Commits on Jul 11, 2023

• Merge pull request rustformers#362 from steventrouble/wte2

  Show description for cf6086c
  philpax

  authored
  cf6086c

Commits on Jul 10, 2023

• Address PR comments

  steventrouble

  committed
  c995ca8

• Explicitly drop model weights

  LLukas22

  committed
  dfaf7ae

• Merge remote-tracking branch 'upstream/main' into feat/cuda-opencl-acceleration

  LLukas22

  committed
  31e3a21

Commits on Jul 9, 2023

• Update gpt2 to use wte if no lm_head

  steventrouble

  committed
  0f841cd

• Fix segfaults due to bad memory view offset.

  steventrouble

  committed
  ff4bb37

• Address PR comments

  steventrouble

  committed
  2badcd9

Commits on Jul 7, 2023

• Add delete tokens test and impl.

  Show description for 2e35b46
  steventrouble

  committed
  2e35b46

• Fix opencl inference

  LLukas22

  committed
  7ec3683

• Merge remote-tracking branch 'upstream/main' into feat/cuda-opencl-acceleration

  LLukas22

  committed
  2a8f62a

Commits on Jul 5, 2023

• fix(llama): use </s> token if available for EOT

  philpax

  committed
  68cefa6

Commits on Jul 2, 2023

• feat(test): check hyperparameters can roundtrip

  philpax

  committed
  00624a4

Commits on Jun 30, 2023

• Update falcon

  LLukas22

  committed
  e1476a4

• Merge remote-tracking branch 'upstream/main' into feat/cuda-opencl-acceleration

  LLukas22

  committed
  b1e8cd0

Commits on Jun 29, 2023

• refactor: vocabulary -> tokenizer

  philpax

  committed
  e7e732e

• refactor: remove unnecessary deps

  philpax

  committed
  ae15cc4

Commits on Jun 28, 2023

• feat(falcon): disable by default

  philpax

  committed
  4716b1d

• Merge branch 'main' into feat/falcon-support

  philpax

  committed
  3b351a8

Commits on Jun 22, 2023

• Made LLama work with CUDA

  LLukas22

  committed
  46fb5e3

Commits on Jun 21, 2023

• Merge branch 'main' into pedal-to-the-metal

  philpax

  committed
  0921b1c

Commits on Jun 20, 2023

• Changed usize to i32 + Scratch buffers

  LLukas22

  committed
  51398ae

Commits on Jun 19, 2023

• metal: move evaluation context to InferenceSession and simplify model building interface

  pixelspark

  committed
  cbbe41c

Commits on Jun 18, 2023

• metal: make mmap work

  pixelspark

  committed
  55a8e55

• metal: make Metal inference work

  pixelspark

  committed
  eba54c3

• metal: load model weights in Metal context

  pixelspark

  committed
  a557b0e

Pagination

PreviousNext