Commit Graph

14 Commits

Author SHA1 Message Date
Angelos Katharopoulos
9f671228cd
Block sparse MM MoEs (#782)
- Adds SwitchLinear
- Adds QuantizedSwitchLinear
2024-05-21 15:58:08 -07:00
Awni Hannun
ee60e2a9d5
Kv cache (#643)
* in place kv_cache

* fix

* fix kv cache size

* partially fix kv cache dtype

* step kv cache

* multiple of step size

* more teests + kv cache

* more kv cache

* udpate all models to use kv cache
2024-05-08 08:18:13 -07:00
Awni Hannun
d3f8e4aee9
Fix argpartition call in Mixtral and other MOES (#676)
* Update mixtral.py

* fix all moes

---------

Co-authored-by: yuhai-china <yuhai.china@gmail.com>
2024-04-12 11:00:56 -07:00
Awni Hannun
b80adbcc3e
DBRX (#628)
* dbrx

* format

* format

* comments

* change scores slightly

* remove inadvertant import
2024-03-28 21:03:53 -07:00
Awni Hannun
b8a348c1b8
Switch to fast RMS/LN Norm (#603)
* use nn.RMSNorm, use sdpa, cleanup

* bump mlx versions

* minor update

* use fast layer norm

* version bump

* update requirement for whisper

* update requirement for gguf
2024-03-23 07:13:51 -07:00
Awni Hannun
e4b19bb9e1
Make attention faster for a some models (#574)
* make attention faster for a couple models

* remove unused generation flags

* add comment on lora

* include text files as well
2024-03-14 21:35:54 -07:00
Awni Hannun
7cdd1b69ac
Enable unit testing in Circle and start some MLX LM tests (#545)
* add a few tests for mlx lm

* add a few tests for mlx lm

* add a few tests for mlx lm

* more tests / cleanup
2024-03-07 09:31:57 -08:00
Awni Hannun
f24edfa9dc
[mlx-lm] Add precompiled normalizations (#451)
* add precompiled normalizations

* nits
2024-02-22 12:40:55 -08:00
Awni Hannun
8fd953ee2b
Support for slerp merging models (#455)
* support for slerp merging models

* docs

* update docs

* format'
2024-02-19 20:37:15 -08:00
devonthomas35
cc671cd1c7
Mixtral: Fix non-default arg follows default exception (#450)
Mixtral models throw the following exception
```
Traceback (most recent call last):
  File "<frozen runpy>", line 198, in _run_module_as_main
  File "<frozen runpy>", line 88, in _run_code
  File "/opt/homebrew/anaconda3/lib/python3.11/site-packages/mlx_lm/generate.py", line 119, in <module>
    main(args)
  File "/opt/homebrew/anaconda3/lib/python3.11/site-packages/mlx_lm/generate.py", line 96, in main
    model, tokenizer = load(args.model, tokenizer_config=tokenizer_config)
                       ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/homebrew/anaconda3/lib/python3.11/site-packages/mlx_lm/utils.py", line 278, in load
    model = load_model(model_path)
            ^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/homebrew/anaconda3/lib/python3.11/site-packages/mlx_lm/utils.py", line 221, in load_model
    model_class, model_args_class = _get_classes(config=config)
                                    ^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/homebrew/anaconda3/lib/python3.11/site-packages/mlx_lm/utils.py", line 46, in _get_classes
    arch = importlib.import_module(f"mlx_lm.models.{model_type}")
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/homebrew/anaconda3/lib/python3.11/importlib/__init__.py", line 126, in import_module
    return _bootstrap._gcd_import(name[level:], package, level)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "<frozen importlib._bootstrap>", line 1204, in _gcd_import
  File "<frozen importlib._bootstrap>", line 1176, in _find_and_load
  File "<frozen importlib._bootstrap>", line 1147, in _find_and_load_unlocked
  File "<frozen importlib._bootstrap>", line 690, in _load_unlocked
  File "<frozen importlib._bootstrap_external>", line 940, in exec_module
  File "<frozen importlib._bootstrap>", line 241, in _call_with_frames_removed
  File "/opt/homebrew/anaconda3/lib/python3.11/site-packages/mlx_lm/models/mixtral.py", line 11, in <module>
    @dataclass
     ^^^^^^^^^
  File "/opt/homebrew/anaconda3/lib/python3.11/dataclasses.py", line 1230, in dataclass
    return wrap(cls)
           ^^^^^^^^^
  File "/opt/homebrew/anaconda3/lib/python3.11/dataclasses.py", line 1220, in wrap
    return _process_class(cls, init, repr, eq, order, unsafe_hash,
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/homebrew/anaconda3/lib/python3.11/dataclasses.py", line 1027, in _process_class
    _init_fn(all_init_fields,
  File "/opt/homebrew/anaconda3/lib/python3.11/dataclasses.py", line 545, in _init_fn
    raise TypeError(f'non-default argument {f.name!r} '
TypeError: non-default argument 'model_type' follows default argument
```
2024-02-18 13:30:26 -08:00
Angelos Katharopoulos
f71e965d57
Change gqa to use repeat instead of concatenate (#443) 2024-02-14 17:40:11 -08:00
Awni Hannun
d4666615bb
Lazy import + refactor Lora layer addition (#426)
* lazy model import in mlx_lm

* change lora loading

* fix olmo lora

* remove a bunch of unused stuff from plamo

* move phixtral to mlx-lm and out of llms/
2024-02-12 10:51:02 -08:00
Anchen
362e88a744
feat: move lora into mlx-lm (#337)
* feat: Add lora and qlora training to mlx-lm


---------

Co-authored-by: Awni Hannun <awni@apple.com>
2024-01-23 08:44:37 -08:00
Anchen
195bec2fa3
feat(mlx_lm): add mixtral support in mlx_lm (#318)
* feat: add mixtral support in mlx_lm

* chore: update doc
2024-01-15 07:18:14 -08:00