wehub-resource-sync
34 lines
1.1 KiB
YAML
34 lines
1.1 KiB
YAML
# Copyright (c) Facebook, Inc. and its affiliates.
|
|
#
|
|
# This source code is licensed under the MIT license found in the
|
|
# LICENSE file in the root directory of this source tree.
|
|
|
|
# This file defines example configuration arguments for quantizing
|
|
# a transformer model with product quantization
|
|
|
|
# Number of Centroids for Product Quantization, by default 256 (byte-aligned)
|
|
n_centroids:
|
|
Linear:
|
|
key: in_features
|
|
value: {"*": 256}
|
|
Embedding:
|
|
key: embedding_dim
|
|
value: {"*": 256}
|
|
|
|
# Block Sizes for Product Quantization
|
|
# We suggest: 8 for FFN, 4 for ATTN, 4 for embedding projections, 8 for embeddings
|
|
block_sizes:
|
|
Linear:
|
|
key: fuzzy_name
|
|
value: {fc: 8, attn: 4, emb: 4}
|
|
Embedding:
|
|
key: fuzzy_name
|
|
value: {emb: 8}
|
|
|
|
# Layers to Quantize Sequentially
|
|
# We suggest: first FFN, then EMB, then ATTN
|
|
layers_to_quantize:
|
|
- decoder\\.layers\\.\d+\\.fc[12]
|
|
- decoder\\.embed_tokens\\.embeddings\\.[012]\\.[01]
|
|
- decoder\\.layers\\.\d+\\.self_attn\\.(k_proj|v_proj|q_proj|out_proj)
|