huggingface/text-generation-inference
5
0
Fork 0
mirror of https://github.com/huggingface/text-generation-inference.git synced 2025-04-21 23:12:07 +00:00
Code Issues Packages Projects Releases Wiki Activity
5d121a9705
text-generation-inference/server/text_generation_server/layers
History
Daniël de Kok 9935720c87
Add support for repacking AWQ weights for GPTQ-Marlin (#2278) 
* Add support for repacking AWQ weights for GPTQ-Marlin

So far we couldn't support AWQ because virtually all AWQ models use
symmetric quantization, which GPTQ-Marlin did not suppors. GPTQ-Marlin
has recently added support AWQ repacking and AWQ asymmetric quantization
(zero_point=True).

This change updates all GPTQ-Marlin kernels from upstream and wires up
AWQ support. For now enabling AWQ using Marlin requires running TGI with
`--quantize gptq`.

* Enable Marlin for supported AWQ configurations by default

This makes the AWQ -> GPTQ repack test redundant, since we are now
testing this with the regular AWQ test.
2024-07-23 13:08:20 +02:00
..
attention Softcapping for gemma2. (#2273) 2024-07-22 18:27:10 +02:00
awq Support AWQ quantization with bias (#2117) 2024-06-25 21:09:00 +02:00
gptq Add support for repacking AWQ weights for GPTQ-Marlin (#2278) 2024-07-23 13:08:20 +02:00
__init__.py Enable multiple LoRa adapters (#2010) 2024-06-25 14:46:27 -04:00
bnb.py feat(fp8): use fbgemm kernels and load fp8 weights directly (#2248) 2024-07-20 19:02:04 +02:00
conv.py Refactor layers. (#1866) 2024-05-13 12:44:30 +02:00
eetq.py feat(fp8): use fbgemm kernels and load fp8 weights directly (#2248) 2024-07-20 19:02:04 +02:00
exl2.py Add support for Deepseek V2 (#2224) 2024-07-19 17:23:20 +02:00
fp8.py fix(l4): fix fp8 logic on l4 (#2277) 2024-07-23 11:24:29 +02:00
layernorm.py Removing IPEX_AVAIL. (#2115) 2024-06-25 13:20:57 +02:00
linear.py Improve the handling of quantized weights (#2250) 2024-07-19 09:37:39 +02:00
lora.py Enable multiple LoRa adapters (#2010) 2024-06-25 14:46:27 -04:00
marlin.py Add support for repacking AWQ weights for GPTQ-Marlin (#2278) 2024-07-23 13:08:20 +02:00
medusa.py fix: use path inside of speculator config (#1935) 2024-05-22 20:46:29 +02:00
mlp.py MLPSpeculator. (#1865) 2024-05-14 12:33:18 +02:00
rotary.py Add support for Deepseek V2 (#2224) 2024-07-19 17:23:20 +02:00
speculative.py MLPSpeculator. (#1865) 2024-05-14 12:33:18 +02:00
tensor_parallel.py Improve the handling of quantized weights (#2250) 2024-07-19 09:37:39 +02:00