[package]
edition = "2021"
name = "ferrum-quantization"
version = "0.7.2"
authors = ["Ferrum Team"]
build = false
autolib = false
autobins = false
autoexamples = false
autotests = false
autobenches = false
description = "Weight-format abstraction (Dense / GPTQ / AWQ / GGUF) for Ferrum models"
readme = "README.md"
license = "MIT"
[features]
cuda = ["ferrum-kernels/cuda"]
default = []
metal = ["ferrum-kernels/metal"]
[lib]
name = "ferrum_quantization"
path = "src/lib.rs"
[[test]]
name = "gguf_linear_test"
path = "tests/gguf_linear_test.rs"
[[test]]
name = "gguf_loader_moe_test"
path = "tests/gguf_loader_moe_test.rs"
[[test]]
name = "gguf_loader_test"
path = "tests/gguf_loader_test.rs"
[[test]]
name = "gguf_parse_test"
path = "tests/gguf_parse_test.rs"
[[test]]
name = "gptq_parity_test"
path = "tests/gptq_parity_test.rs"
[dependencies.candle-core]
version = "0.9.2"
[dependencies.ferrum-kernels]
version = "0.7.2"
[dependencies.ferrum-types]
version = "0.7.2"
[dependencies.half]
version = "2.5.0"
features = [
"num-traits",
"use-intrinsics",
"rand_distr",
]
[dependencies.memmap2]
version = "0.9"
[dependencies.safetensors]
version = "0.4"
[dependencies.serde]
version = "1.0"
features = ["derive"]
[dependencies.serde_json]
version = "1.0"
[dependencies.tracing]
version = "0.1"
[dev-dependencies.tempfile]
version = "3.8"