Model · alpaca-7b

Alpaca 7B

Developer: Stanford CRFM (Tatsu Lab)
Availability: available · checked 2026-09-24 · source
Released by the developer as a weight diff, to be applied to LLaMA 7B weights.

Raw record: /data/models/alpaca-7b.json

Fields

id

alpaca-7b

identifiers

huggingface

tatsu-lab/alpaca-7b-wdiff

developer

Stanford CRFM (Tatsu Lab)

release_date

2023-03-13recorded · source

weights_status

open

availability

available · checked 2026-09-24 · source

note:  Released by the developer as a weight diff, to be applied to LLaMA 7B weights.

license

Research-only; commercial use explicitly prohibited. Fine-tuned weights not directly redistributed (LLaMA's own license restricts this) -- release was a training recipe / diff, not confirmed as full weight redistribution.partial · source

architecture

family

decoder_only

note

Inherits LLaMA 7B's architecture unchanged; no separate config published for the fine-tuned model.

n_layers

nullnot_recorded

hidden_size

nullnot_recorded

n_heads

nullnot_recorded

vocab_size

nullnot_recorded

positional_encoding

nullnot_recorded

training_data

52K instruction-following examples generated self-instruct style from OpenAI's text-davinci-003.recorded · source

techniques

instruction-tuning

self-instruct

primary_sources

https://crfm.stanford.edu/2023/03/13/alpaca.html

https://github.com/tatsu-lab/stanford_alpaca

record_history

date:2026-09-24 · change:created from primary sources (Stanford CRFM blog post, GitHub repo) · by:wilson-pruitt + claude ·

date:2026-09-24 · change:availability checked and recorded · by:wilson-pruitt + claude ·

Parents

Weights descend

• fine_tuned_from → LLaMA 7B declared source Stanford CRFM: "fine-tuned from Meta's LLaMA 7B model"

Training data

• trained_on → Alpaca instruction data (52K) declared source Fine-tuning data.

Children

No edges recorded.

Read in

No station on the reading path has touched this record yet.