George Hotz b5a9465b13 llm: add support for moonlight (deepseek MLA) (#15466) ...

* add gguf Q5_0

* it works

* rebase

* simpler test

* class

* less diff

* dicts

* normal names

* simplify

* this

* simpler

* work

* work

2026-04-11 10:32:48 +08:00

..

__init__.py

move files that pass with NULL=1 to test/null (#14508)

2026-02-03 13:52:36 +08:00

test_attention.py

jit: graphing in uops (#15489)

2026-03-27 19:09:02 +03:00

test_autogen.py

c.Struct cleanup (#15640)

2026-04-08 20:07:16 -04:00

test_compile_failures.py

DEV has arch (#15577)

2026-04-03 19:17:19 -04:00

test_const_folding.py

fixes and test relaxations from prealloc_bufs (#14875)

2026-02-19 11:37:25 +08:00

test_device.py

move HCQ_VISIBLE_DEVICES into DEV (#15668)

2026-04-09 22:01:35 -04:00

test_disk_cache.py

move files that pass with NULL=1 to test/null (#14508)

2026-02-03 13:52:36 +08:00

test_dtype_spec.py

add weakint to promo_lattice (#15463)

2026-03-25 00:27:34 -04:00

test_dtype.py

move files that pass with NULL=1 to test/null (#14508)

2026-02-03 13:52:36 +08:00

test_elf.py

move files that pass with NULL=1 to test/null (#14508)

2026-02-03 13:52:36 +08:00

test_gc.py

move more to null/unit tests (#14658)

2026-02-10 13:35:17 +08:00

test_gpudims.py

DEV has arch (#15577)

2026-04-03 19:17:19 -04:00

test_gradient.py

move more tests to test/null, split some existing ones (#14512)

2026-02-03 20:20:20 +08:00

test_graph_rewrite.py

dtypes.index -> dtypes.weakint (#15377)

2026-03-20 01:08:46 -04:00

test_hcq_iface.py

move files that pass with NULL=1 to test/null (#14508)

2026-02-03 13:52:36 +08:00

test_helpers.py

move more tests to test/null, split some existing ones (#14512)

2026-02-03 20:20:20 +08:00

test_indexing.py

clean up test/null/test_indexing.py (#14720)

2026-02-12 22:36:53 -05:00

test_linearizer_failures.py

dtypes.index -> dtypes.weakint (#15377)

2026-03-20 01:08:46 -04:00

test_linearizer_rewrite.py

KernelInfo is required on get_program (#14571)

2026-02-06 10:49:27 +08:00

test_llm_server.py

llm: add support for moonlight (deepseek MLA) (#15466)

2026-04-11 10:32:48 +08:00

test_llm_tokenizer.py

llm: fix streaming UTF-8 decode (#15653)

2026-04-10 17:01:02 +08:00

test_memory_planner.py

memplan on linears (#15422)

2026-03-23 19:50:16 +08:00

test_method_cache.py

move files that pass with NULL=1 to test/null (#14508)

2026-02-03 13:52:36 +08:00

test_microbenchmarks.py

move files that pass with NULL=1 to test/null (#14508)

2026-02-03 13:52:36 +08:00

test_mnist_dataset.py

buffer view is like buffer (#15082)

2026-03-03 09:52:33 +08:00

test_multitensor.py

memplanner opt copy bufs (#15110)

2026-03-08 22:28:01 +03:00

test_null.py

DEV has arch (#15577)

2026-04-03 19:17:19 -04:00

test_pattern_matcher.py

dtype mixin (#14763)

2026-02-15 16:03:48 +08:00

test_process_replay.py

fix opt in process replay [pr] (#14599)

2026-02-06 16:49:56 -05:00

test_real_world.py

memplan on linears (#15422)

2026-03-23 19:50:16 +08:00

test_rearrange_einops.py

move more tests to test/null, split some existing ones (#14512)

2026-02-03 20:20:20 +08:00

test_resnet.py

everything has dtype.long now (#14661)

2026-02-10 15:08:50 +08:00

test_rewrite_bottom_up_gate.py

move files that pass with NULL=1 to test/null (#14508)

2026-02-03 13:52:36 +08:00

test_schedule_cache.py

move more tests to test/null, split some existing ones (#14512)

2026-02-03 20:20:20 +08:00

test_schedule.py

improve ReduceMixin.all (#15544)

2026-03-31 07:54:27 -04:00

test_simplify_valid_idx.py

few more fold valid tests (#15509)

2026-03-27 10:38:42 -04:00

test_symbolic_failures.py

move files that pass with NULL=1 to test/null (#14508)

2026-02-03 13:52:36 +08:00

test_symbolic_tensor.py

move more tests to test/null, split some existing ones (#14512)

2026-02-03 20:20:20 +08:00

test_tensor_io.py

move more tests to test/null, split some existing ones (#14512)

2026-02-03 20:20:20 +08:00

test_tensor_metadata.py

jit: captures linears, not execitems (#15399)

2026-03-21 16:32:12 +08:00

test_tensor_uop_representation.py

more test relaxations from prealloc_bufs (#14880)

2026-02-19 14:23:28 +08:00

test_tensor.py

simple CreationMixin (#15567)

2026-04-01 23:00:56 -04:00

test_tinyfs.py

revert some late_buffer_view change (#14578)

2026-02-05 22:51:40 -05:00

test_tqdm.py

relax test_tqdm_perf (#15134)

2026-03-04 12:58:47 -05:00

test_transcendental_helpers.py

rename all DEFINE_GLOBAL to PARAM (#14511)

2026-02-03 15:09:38 +08:00

test_uop_graph.py

shared _normalize_indices for getitem (#15625)

2026-04-06 17:45:36 -04:00

test_uop_repr.py

move files that pass with NULL=1 to test/null (#14508)

2026-02-03 13:52:36 +08:00

test_uop_resolve.py

move files that pass with NULL=1 to test/null (#14508)

2026-02-03 13:52:36 +08:00

test_uop_symbolic.py

support int operand for shifts (#15618)

2026-04-06 12:32:12 -04:00

test_uop_vmin_vmax.py

XOR simplifcation rules (#15512)

2026-03-27 23:23:27 -04:00

test_uops_stats.py

DEV has arch (#15577)

2026-04-03 19:17:19 -04:00

test_uops.py

fixes and test relaxations from prealloc_bufs (#14875)

2026-02-19 11:37:25 +08:00

test_upat_compile.py

move files that pass with NULL=1 to test/null (#14508)

2026-02-03 13:52:36 +08:00

test_validate_oob.py

dtypes.index -> dtypes.weakint (#15377)

2026-03-20 01:08:46 -04:00

test_viz.py

viz: add # of bytes to sdma (#15639)

2026-04-08 17:43:37 +09:00

test_winograd.py

preallocate all realized buffers (#14823)

2026-02-20 20:05:54 +08:00