kyegomez / zeta

Build high-performance AI models with modular building blocks

https://zeta.apac.ai

kyegomez/zeta Issues

[BUG] zeta/nn/modules/multiclass_label.py is empty
Closed 9 days ago1
[BUG] VisionMambaBlock example
Updated 19 days ago3
[BUG] no default value for depth of AttentionLayer
Closed a month ago1
[BUG] Attend , dim, MaxVit
Closed a month ago1
[BUG] RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cuda:0 and cpu! (when checking argument for argument weight in method wrapper_CUDA__native_layer_norm)
Updated a month ago
[BUG] RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cuda:0 and cpu!
Updated a month ago4
[BUG] Install errors
Updated a month ago2
How to check Zeta Version installed? zeta.__version__ does not seem to work
Closed 2 months ago3
[BUG] RMSNorm Implementation
Updated 2 months ago1
[BUG] [DOCS] XPOS
Closed 2 months ago2
[BUG] [DOCS] missing/misplaced docs
Closed 2 months ago2
[BUG] [DPO - Direct Policy Optimization][RuntimeError: mat1 and mat2 must have the same dtype, but got Long and Float]
Closed 2 months ago1
[BUG] [DOCS] SinusoidalEmbeddings rotate_half
Closed 2 months ago1
[BUG] SinusoidalEmbedding rotate_every_two , related to #106
Closed 2 months ago1
[TEST] test_cast_tuple.py - failing test - suggest removing
Closed 2 months ago2
[BUG] [DOCS] TruncatedRotaryEmbedding example KeyError: "attribute 'inv_freq' already exists"
Closed 2 months ago1
[BUG] [DOCS] PositionalEmbedding example IndexError: Dimension out of range
Closed 2 months ago1
[TEST] cloud/test_main failures
Closed 2 months ago1
[TEST] models/test_andromeda num_tokens attribute
Closed 2 months ago1
[BUG] how could i find a zata version that compatiable with python3.8 and torch 1.12
Closed 2 months ago2
[BUG] utils/main/pad_at_dim - recommend refactoring to use torch.nn.functional.pad
Closed 2 months ago1
Relationship with x-transformers
Closed 2 months ago1
[BUG] Running import zeta immediately fails with "RuntimeError: mat1 and mat2 shapes cannot be multiplied (512x4 and 512x512)"
Closed 3 months ago3
[BUG] [DOCS] Custom MLP - Forward Pass
Closed 3 months ago1
[BUG] Pip should be updated
Closed 3 months ago3
[BUG] [DOCS] MBConv function
Closed 3 months ago
[BUG]MambaBlock no working on MacOS, even just import
Closed 4 months ago5
[BUG] [DOCS] VitTransformerBlock examples have einops error
Closed 4 months ago2
[BUG] [DOCS] RelativePositionBias calling n_heads sb num_heads
Closed 4 months ago1
[BUG] [DOCS] [IMPORT] example fix - import log_prob from DPO
Closed 4 months ago1
[BUG] [DOCS] [IMPORT] XPOS
Closed 4 months ago2
[BUGF] spelling SpacialTransformer should be SpatialTransformer
Closed 4 months ago1
[BUG] VisionEmbedding readme example error
Closed 4 months ago2
[TEST] adaptive_rmsnorm - no module named
Closed 4 months ago5
[TEST] half_bit_linear - no module named
Closed 4 months ago5
[BUG] FusedDenseGELUDense, FusedDropoutLayerNorm not exported in zeta/nn/__init__.py
Closed 4 months ago1
[CONF] generator-generic-ossf-slsa3-publish.yml - needs initial config
Closed 4 months ago
[TEST]
Closed 4 months ago1
[BUG] ERROR: (gcloud.services.enable) FAILED_PRECONDITION:
Closed 5 months ago1
[BUG] tensorflow imported
Closed 5 months ago
[BUG] TypeError: new(): invalid data type 'str'
Closed 5 months ago
[BUG] test_bitlinear duplicated.
Closed 5 months ago
[BUG] remove __pycache__ / .pyc files and/or use a unique basename for your test file modules
Closed 5 months ago
[BUG] ERROR collecting rl/test_prioritizedreplybuffer.py - ModuleNotFoundError: No module named 'sumtree'
Closed 5 months ago
[BUG] ERROR collecting rl/test_prioritizedsequencereplybuffer.py - ModuleNotFoundError: No module named 'sumtree'
Closed 5 months ago1
[BUG] ModuleNotFoundError: No module named 'zeta.cloud'
Closed 5 months ago1
[BUG] no module named zeta
Closed 5 months ago1
[BUG] pip apex is not nvidia apex
Closed 5 months ago
[BUG] ModuleNotFoundError: No module named 'zeta.structs.attn_layers'
Closed 5 months ago2
[BUG] fix main.py -> test_main.py
Closed 5 months ago