Sync: tinygrad/tinygrad:master → sunnypilot/tinygrad:master#6
Merged
Conversation
* llm: glm 4.5 air * clean * clean * remove gguf_size
don't use UOp.ufix for const_like, it will broadcast dtype soon
index arg should have weakint dtype
* cuda graph to linear * fix * keep as old for now * x * x
* x * jit: precompile as sep step * x * s * x * x * x * ? * ? * x * x * viz * f * x * u * x * x
* UOp binary op broadcasts dtype matches Tensor * fix * fix?
instead of implicit `to`. this matches torch
* viz/cli: accept one path argument * -t default * hm * only the -t change
* sqtt: update wave end packet names * update wavestart and emu
* metal to linear * x * x * fix
use `.eq` instead of `==`
…#15902) also use `.ne` to fix `# type: ignore[comparison-overlap]`
* avg_pool2d, max_pool2d to mixin * fix * just dtype * that
* hcq * f * f * linter
* update nir renderer for new image style * don't cast image indexes
* ci: add IMAGE=1 compile-only tests * fix
all the slowness is scheduling
* llama: tie store to grad directly * llama: set mp flags * llama: non fused grad fp8 quantize path
* llm: precise device * llm: pass device to precompute_freqs_cis
remove some no-op slices
* keep this for now * early graph
same kernels
* add gelu approximate="none" (match pytorch) * lint * pass through onnx Gelu approximate * type annotate * explicit math.sqrt * keep tinygrad's gelu approximate="tanh" default
* viz/cli: readable uop json repr * work * better
* disable fast idiv by default, it's broken * fix fast idiv tests
* share more with graph * comment
* add implicit broadcasting to shape * NOOP/ALLREDUCE fixes
backward through clone crashes now
also decouple from requires_grad
final goal is to remove requires_grad, first change the default to True, and don't allow None
|
This branch currently is behind tinygrad/master. The line count difference bot is disabled. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
20 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
No description provided.