Implement GPTQ quantization

[EricLBuehler]

This PR adds GPTQ quantization (paper here) support.

Refs: #418, #448.

[github-actions]

Code Metrics Report

  ===============================================================================
 Language            Files        Lines         Code     Comments       Blanks
===============================================================================
 C Header                2           35           28            0            7
 Dockerfile              1           34           25            0            9
 Happy                   1          442          369            0           73
 JSON                   11          102          101            0            1
 Python                 45         1993         1695           62          236
 TOML                   19          574          506           11           57
-------------------------------------------------------------------------------
 Jupyter Notebooks       4            0            0            0            0
 |- Markdown             2           77           32           31           14
 |- Python               2          196          169            1           26
 (Total)                            273          201           32           40
-------------------------------------------------------------------------------
 Markdown               25         1842            0         1385          457
 |- BASH                 5          101           98            0            3
 |- JSON                 1           12           12            0            0
 |- Python               5           92           82            0           10
 |- Rust                 6          408          365           19           24
 |- TOML                 2           75           63            0           12
 (Total)                           2530          620         1404          506
-------------------------------------------------------------------------------
 Rust                  173        55983        50835         1000         4148
 |- Markdown            92          864           13          801           50
 (Total)                          56847        50848         1801         4198
===============================================================================
 Total                 282        61005        53559         2458         4988
===============================================================================
  

[EricLBuehler]

cargo run --features cuda -- -i plain -m kaitchup/Phi-3-mini-4k-instruct-gptq-4bit -a phi3

[BuildBackBuehler]

Broke my heart when I went to try Mistral_Large 2-bit EQAT (AutoGPTQ) on my M1 and only then saw no Mac support 😭. Wondering when might that come around? If adding that support for MPS/Metal was not too high-level expert knowledge prereq'd of a task I wouldn't mind taking a swing at it 😂

[EricLBuehler]

@BuildBackBuehler If you could add this, it would be amazing!

I haven't seen GPTQ kernels on Mac though, if you can find any it shouldn't be too hard to add it and I would appreciate it if you take a shot!