FacebookResearch LLaMA: Open and Efficient Foundation Language Models

  • CategoryApps
  • TypeOther
  • LanguageEnglish
  • Total size219 GB
  • Uploaded ByMybugana
  • Downloads19
  • Last checkedMar. 09th '23
  • Date uploadedMar. 08th '23
  • Seeders 1
  • Leechers5

Infohash : E6CC827D48F38434FF57E3F3ACE62C5CC65ACB1B

Quote:

We introduce LLaMA, a collection of foundation language models ranging from 7B to 65B parameters. We train our models on trillions of tokens, and show that it is possible to train state-of-the-art models using publicly available datasets exclusively, without resorting to proprietary and inaccessible datasets. In particular, LLaMA-13B outperforms GPT-3 (175B) on most benchmarks, and LLaMA-65B is competitive with the best models, Chinchilla70B and PaLM-540B. We release all our models to the research community.

Files:

LLaMA 65B
  • consolidated.00.pth (15.2 GB)
  • checklist.chk (0.5 KB)
  • consolidated.01.pth (15.2 GB)
  • consolidated.02.pth (15.2 GB)
  • consolidated.03.pth (15.2 GB)
  • consolidated.04.pth (15.2 GB)
  • consolidated.05.pth (15.2 GB)
  • consolidated.06.pth (15.2 GB)
  • consolidated.07.pth (15.2 GB)
  • params.json (0.1 KB)
  • tokenizer_checklist.chk (0.0 KB)
  • 7B
    • checklist.chk (0.1 KB)
    • consolidated.00.pth (12.6 GB)
    • params.json (0.1 KB)
    13B
    • checklist.chk (0.2 KB)
    • consolidated.00.pth (12.1 GB)
    • consolidated.01.pth (12.1 GB)
    • params.json (0.1 KB)
    30B
    • checklist.chk (0.3 KB)
    • consolidated.00.pth (15.1 GB)
    • consolidated.01.pth (15.1 GB)
    • consolidated.02.pth (15.1 GB)
    • consolidated.03.pth (15.1 GB)
    • params.json (0.1 KB)
  • tokenizer.model (488.0 KB)

Code:

  • udp://tracker.leechers-paradise.org:6969/announce