* write python emulator from RDNA3 psuedocode in pdf
* emu2
* more emu
* working
* more psueod
* progress
* cleanups
* delete junk
* delete stale files
* just emu
* work
* emu compare
* bemu
* cleanups and more failures
* revert bench emu
* fix emu cmp
* four tests fail
* bugfixes
* dsl
* ext
* refactor
* dsl
* div scale fix
* test_emu
* fix emu tests
* pcode
* test pcode
* top imports
* fix test_emu to use run_asm
* emu tests on real hardware
* more tests
* more emu tests
* more
* work
* work
* bug fix
* bugfixes
* fix fp16 gemm
* all ops tests pass in emulator
* fix llvm tests
* fix a few more tests
* fix mockgpu timeout
`UOp.render()` crashed with `IndexError: tuple index out of range` when
the UOp graph contained a `VECTORIZE` with empty `src=()`. This occurs
when reshaping to scalar shape `()`, e.g., `Tensor.ones(4).sum()`.
The bug was in the renderer's VECTORIZE pattern: `all_same(())` returns
`True` (vacuous truth), causing the code to access `x.src[0]` on an
empty tuple.
- Fix `IndexError` when calling `UOp.render()` on graphs containing
empty `VECTORIZE` nodes.
- Add test for empty `VECTORIZE` rendering.
* having fun with python asm dsl
* rdna3
* meh
* all in rdna3
* work
* more work
* work
* integration
* tests
* simpler
* simpler
* asm
* better
* simpler
* progress
* emu
* simpler
* emu
* tests
* types
* vopd
* cleaups
* work
* memory ranges
* add tracing
* refactors
* run_asm exit
* more readable
* compare to remu
* test gemm
* bug + stale
* more tests
* refactor
* tests fix
* more ins
* more instructions
* refactor
* faster
* match case
* match case
* simpler
* work
* tests
* run_asm
* work
* bug fixes
* more emu
* alu/emu
* refactor
* no pipeline emu yet
* alu direct
* fix
* bugfixes + new test
* fix exceptions in emulators
* update gen.py
* pylint
* no pdf
* improve bench_emu
* speedups
* cleanups
* more tests
* remove ExecItem and merge it with ScheduleItem
* less diff
* fix issues
* min diff
* don't change bufs in _lower
* min diff
* update
* revert
* fixes
* diff
* support symbolic shapes in split/chunk when split dim is concrete
Previously split() and chunk() required all dimensions to be concrete.
Now they only require the dimension being split to be concrete, allowing
them to work with tensors that have symbolic shapes in other dimensions.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
* update CLAUDE.md: add pre-commit and no-amend rules
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
* fix dim resolution order in split/chunk
Ensure dim_sz is retrieved after dim is resolved, not before.
The previous one-liner evaluated self.shape[dim] with the original
unresolved dim value.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>
* qwen model is mostly working
* add Q4_K quantization support to GGUF parser, add qwen3:1.7b model
- Add Q4_K (type 12) dequantization in nn/state.py
- Add qwen3:1.7b model using Q4_K_M quantization (smaller than Q8_0)
- Make bos_token_id optional for models like Qwen3 that don't have it
- Fix line length issues and add preset parameter to SimpleTokenizer
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
* smaller diff
* test dequant
* half split
* better
* simple tok
* mock token
* polish
* better
* fix
* replace
---------
Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>
* schedule: cache unbinds for consistent cache keys
strip BIND values before computing cache key so different bound values
(e.g. KV cache positions) hit the same schedule cache entry.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
* spec: allow single-src BIND for schedule cache key normalization
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
* docs: add lessons learned to CLAUDE.md
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
* more claude.md
---------
Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>
* schedule: cache unbinds for consistent cache keys
different bound variable values (e.g. kv cache positions) now produce
the same schedule cache key by unbinding BIND(DEFINE_VAR, CONST) before
computing the cache key and rebinding after lookup.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
* schedule: cache unbinds for consistent cache keys
When scheduling, BIND(DEFINE_VAR, CONST) nodes are now unbound to
tagged DEFINE_VARs before computing the cache key. This ensures that
the same computation with different bound values (e.g., different
KV cache positions in LLM) gets the same cache key and reuses the
cached schedule.
The fix:
- pm_pre_sched_cache: replaces BIND with tagged DEFINE_VAR
- pm_post_sched_cache: restores tagged DEFINE_VAR back to original BIND
- pm_remove_rangeify_tags: excludes DEFINE_VAR to preserve tags through rangeify
- var_vals extracted from BINDs before cache key computation
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
* schedule: fix BIND handling and add CLAUDE.md
- Handle BIND to RANGE in create_schedule (not matched by CONST pattern)
- Assert all BINDs on same variable have same value
- Add CLAUDE.md codebase guide
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>
