Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Dec 13, 2024

• improve type hinting and fix use_cache (#680)

  casper-hansen

  authored
  f1abb8e

Commits on Dec 6, 2024

• bump to 0.2.7.post3 (#674)

  casper-hansen

  authored
  cbd6a75

• pin huggingface_hub (#673)

  casper-hansen

  authored
  b0ac9f3

• install hub main branch (#672)

  casper-hansen

  authored
  7147682

• Fix missing embed_tokens (#671)

  casper-hansen

  authored
  8c5dcc6

Commits on Dec 3, 2024

• multi-gpu fix (#668)

  casper-hansen

  authored
  f2171f3

Commits on Dec 2, 2024

• fix "Expected all tensors to be on the same device" (#664)

  casper-hansen

  authored
  9f13358

Commits on Nov 27, 2024

• fix exaone import (#659)

  casper-hansen

  authored
  0d1906f

Commits on Nov 26, 2024

• Add EXAONE support (#651)

  lgai-exaone

  authored
  4b10b31

• Fused attention: Switch to Flash Decoding (#656)

  casper-hansen

  authored
  dfe396a

Commits on Nov 18, 2024

• Post release 2 - All additional packages goes into extras (#653)

  casper-hansen

  authored
  167c780

Commits on Nov 16, 2024

• post release 1 (#650)

  casper-hansen

  authored
  947a839

• Minimum of torch 2.2.0 during build (#649)

  casper-hansen

  authored
  8f71dc8

• fix download wheels script

  casper-hansen

  committed
  887d08b

• Only build once (#648)

  casper-hansen

  authored
  39f6334

• New release (0.2.7) + Fix build (#647)

  casper-hansen

  authored
  3b0ae79

Commits on Nov 14, 2024

• Replace custom sharding with save_torch_state_dict from huggingface_hub (#644)

  casper-hansen

  authored
  a28c747

• Enable Intel GPU path and lora finetune and change examples to support different devices (#631)

  Show description for 419a242
  

  jiqing-feng

  and

  casper-hansen

  authored
  419a242

• support minicpm3.0 (#605)

  Show description for b42e3c3
  

  LDLINGLINGLING

  and

  casper-hansen

  authored
  b42e3c3

• Add support for Phi-3-vision series model (#596)

  Show description for 0187ac1
  

  Isotr0py

  and

  casper-hansen

  authored
  0187ac1

• add qwen2vl support (#599)

  Show description for 76bc0a8
  

  kq-chen

  and

  casper-hansen

  authored
  76bc0a8

• fix for "two devices" issue due to RoPE changes (#630)

  davedgd

  authored
  12c91b7

Commits on Sep 13, 2024

• enable awq ipex linear in transformers (#610)

  jiqing-feng

  authored
  7954766

Commits on Sep 12, 2024

• Replace itrex qbits to ipex woq linear (#549)

  Show description for eab1a4a
  

  jiqing-feng

  and

  casper-hansen

  authored
  eab1a4a

• Let installed PyTorch decide required version number (#573)

  Show description for d5ec43d
  

  wasertech

  and

  casper-hansen

  authored
  d5ec43d

• device_map defaults to auto (#607)

  casper-hansen

  authored
  1523880

• AWQ Triton kernels. Make autoawq-kernels optional. (#608)

  casper-hansen

  authored
  ae77736

Commits on Aug 28, 2024

• Implement NO_KERNELS flag and update torch requirement (#582)

  Show description for 8d903b2
  devin-ai-integration[bot]

  authored
  8d903b2

• doc: replace a broken example with a working one (#595)

  stas00

  authored
  79258d6

• quantization fails with old datasets (#593)

  stas00

  authored
  97f4a77

Commits on Aug 12, 2024

• Add Internlm2 support (#576)

  Show description for 6f14fc7
  

  Crystalcareai

  and

  casper-hansen

  authored
  6f14fc7

Commits on Aug 4, 2024

• ignore onnx in ignore_patterns (#570)

  casper-hansen

  authored
  4d5c4a9

• Add Gemma2 support. (#562)

  radi-cho

  authored
  b9e9b73

• fix: pass rope_theta argument when initializing LlamaLikeBlock for models like qwen2, mistral, etc. (#568)

  Show description for 202b967
  

  Shuai-Xie

  and

  Shuai Xie

  authored
  202b967

Commits on Jul 24, 2024

• Update script for Mistral Large

  casper-hansen

  committed
  2e2635b

Pagination

PreviousNext