Daniël de Kok 4594e6faba Add support for Marlin-quantized models ... This change adds support for Marlin-quantized models. Marlin is an FP16xINT4 matmul kernel, which provides good speedups decoding batches of 16-32 tokens. It supports quantized models with symmetric quantization, groupsize -1 or 128, and 4-bit. Tested with: - Llama 2 - Llama 3 - Phi 3		2024-06-06 13:16:52 +02:00
..
__init__.py	feat(server): Add native support for PEFT Lora models (#762)	2023-08-03 17:22:45 +02:00
convert.py	Force weights_only (before fully breaking pickle files anyway). (#1710)	2024-04-05 19:23:57 +02:00
dist.py	add intel xpu support for TGI (#1475)	2024-04-26 15:48:58 +02:00
hub.py	Fixing the download strategy for ibm-fms (#1917)	2024-05-18 13:31:24 +02:00
import_utils.py	Purely refactors paged/attention into `layers/attention` and make hardware differences more obvious with 1 file per hardware. (#1986)	2024-05-31 17:57:01 +02:00
log.py	v1.3.4	2023-12-22 15:46:04 +01:00
logits_process.py	Fixing frequency penalty (#1811)	2024-04-30 12:13:23 +02:00
peft.py	fix: fix local loading for .bin models (#1419)	2024-01-09 15:21:00 +01:00
speculate.py	chore: formatting	2023-12-11 14:49:52 +01:00
tokens.py	Use the generation config. (#1808)	2024-04-25 19:41:50 +02:00
watermark.py	Fixing watermark. (#851)	2023-08-16 07:17:26 +02:00
weights.py	Add support for Marlin-quantized models	2024-06-06 13:16:52 +02:00