Skip to content

Actions: bitsandbytes-foundation/bitsandbytes

Actions

Python package

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
python-package.yml
1,064 workflow runs
1,064 workflow runs

Filter by Event

⌥ + click/return to exclude

Filter by Status

⌥ + click/return to exclude

Filter by Branch

⌥ + click/return to exclude

Filter by Actor

⌥ + click/return to exclude
Support skip_zeros in the XPU Triton optimizer kernels Python package #3016: Pull request #2097 synchronize by jiqing-feng
Action required jiqing-feng:xpu-triton-skip-zeros
jiqing-feng:xpu-triton-skip-zeros
Action required
Support skip_zeros in the XPU Triton optimizer kernels Python package #3015: Pull request #2097 opened by jiqing-feng
Action required jiqing-feng:xpu-triton-skip-zeros
jiqing-feng:xpu-triton-skip-zeros
Action required
fix: wrong absmax indexing in AVX512 dequantizeBlockwise4bitCpu Python package #3014: Pull request #2096 opened by NanoCode012
Action required NanoCode012:patch-1
NanoCode012:patch-1
Action required
MPS: only use the Hub GEMV kernel when K is a multiple of blocksize Python package #3013: Pull request #2094 opened by egeozkoc
Action required egeozkoc:fix/mps-hub-gemv-unaligned-k
egeozkoc:fix/mps-hub-gemv-unaligned-k
Action required
test: align post-reload max_error_count with pre-reload check in test_optimizer32bit Python package #3012: Pull request #2093 opened by CS-liujf
Action required CS-liujf:fix/lion-reload-threshold
CS-liujf:fix/lion-reload-threshold
Action required
Fix device parametrization for optimizer benchmarks Python package #3011: Pull request #2092 opened by emme1t
Action required emme1t:fix/optimizer-benchmark-device
emme1t:fix/optimizer-benchmark-device
Action required
fix: pass backend="cpu" when requesting CPU gemm_4bit_forward kernel Python package #3010: Pull request #2090 opened by SIDDARTHAREDDY8
Action required SIDDARTHAREDDY8:bhai-blitz-8-bitsandbytes-1
SIDDARTHAREDDY8:bhai-blitz-8-bitsandbytes-1
Action required
[ROCm] Preserve float bits in SIMT DPP reduction Python package #3009: Pull request #2089 opened by liminfei-amd
Action required liminfei-amd:fix/2087-gfx1100-dpp-bitcast
liminfei-amd:fix/2087-gfx1100-dpp-bitcast
Action required
fix(optim): reject bias_correction and adam_w_mode in LAMB Python package #3008: Pull request #2088 opened by ConnorMoss02
Action required ConnorMoss02:fix/lamb-reject-unsupported-args
ConnorMoss02:fix/lamb-reject-unsupported-args
Action required
fix(optim): reject nesterov and dampening in SGD and LARS Python package #3007: Pull request #2085 opened by ConnorMoss02
Action required ConnorMoss02:fix/sgd-lars-reject-unsupported-nesterov-dampening
ConnorMoss02:fix/sgd-lars-reject-unsupported-nesterov-dampening
Action required
fix(cpu): correctly handle 1D output shape in dequantize_4bit Python package #3006: Pull request #2055 synchronize by SparshM8
1h 19m 18s SparshM8:fix/cpu-dequantize-1d-shape
SparshM8:fix/cpu-dequantize-1d-shape
1h 19m 18s
Fix Params4bit.to() leaving CPU-packed weights packed Python package #3005: Pull request #2083 opened by 22elix3r
Action required 22elix3r:fix/params4bit-to-cpu-packing
22elix3r:fix/params4bit-to-cpu-packing
Action required
Fix PytorchLARS updates when momentum is zero Python package #3004: Pull request #2082 opened by mottopanikeiku
Action required mottopanikeiku:fix/issue-2079
mottopanikeiku:fix/issue-2079
Action required
Warn once when Linear4bit computes in float32 on GPU Python package #3003: Pull request #2077 opened by thangdangjp
Action required thangdangjp:feature/warn-fp32-compute-dtype
thangdangjp:feature/warn-fp32-compute-dtype
Action required
Make AdEMAMix32bit inherit AdEMAMix so it allocates the double state buffer Python package #3002: Pull request #2076 opened by caiotheodoro
Action required caiotheodoro:ademamix32bit-double-buffer
caiotheodoro:ademamix32bit-double-buffer
Action required
chore(deps): bump softprops/action-gh-release from 3.0.2 to 3.0.3 in the actions group Python package #3001: Pull request #2074 opened by dependabot Bot
10m 35s dependabot/github_actions/actions-6512b1d693
dependabot/github_actions/actions-6512b1d693
10m 35s
chore(deps): bump softprops/action-gh-release in the actions group Python package #3000: Commit 832d3c1 pushed by dependabot Bot
9m 28s dependabot/github_actions/actions-6512b1d693
dependabot/github_actions/actions-6512b1d693
9m 28s
[ROCm] Allow explicit GPU architecture override Python package #2999: Pull request #2073 opened by LOGO127
30d 0h 2m 55s LOGO127:fix/rocm-arch-env-override
LOGO127:fix/rocm-arch-env-override
30d 0h 2m 55s
Add AVX2 path for CPU 4-bit (NF4/FP4) blockwise dequantization Python package #2998: Pull request #2072 synchronize by rachitj27
30d 0h 2m 54s rachitj27:feature/cpu-avx2-4bit-dequant
rachitj27:feature/cpu-avx2-4bit-dequant
30d 0h 2m 54s
Add AVX2 path for CPU 4-bit (NF4/FP4) blockwise dequantization Python package #2997: Pull request #2072 opened by rachitj27
30d 0h 2m 54s rachitj27:feature/cpu-avx2-4bit-dequant
rachitj27:feature/cpu-avx2-4bit-dequant
30d 0h 2m 54s
Fix CPU Hub kernel backend selection Python package #2996: Pull request #2065 synchronize by SiluPanda
12m 24s SiluPanda:fix/issue-2064
SiluPanda:fix/issue-2064
12m 24s
fix(cuda): chunk 4-bit quant/dequant launches above INT32_MAX elements (#1785) Python package #2995: Pull request #2071 opened by Anai-Guo
30d 0h 3m 3s Anai-Guo:fix/4bit-chunk-int32-launch
Anai-Guo:fix/4bit-chunk-int32-launch
30d 0h 3m 3s
Feat/mps gemm 4bit bf16 Python package #2994: Pull request #2070 opened by eaglstun
12m 41s eaglstun:feat/mps-gemm-4bit-bf16
eaglstun:feat/mps-gemm-4bit-bf16
12m 41s
fix(optim): fall back to non-paged state on all devices without managed memory Python package #2993: Pull request #2069 synchronize by egeozkoc
30d 0h 1m 41s egeozkoc:fix/paged-optimizer-non-cuda
egeozkoc:fix/paged-optimizer-non-cuda
30d 0h 1m 41s
fix(optim): fall back to non-paged state on all devices without managed memory Python package #2992: Pull request #2069 opened by egeozkoc
9m 16s egeozkoc:fix/paged-optimizer-non-cuda
egeozkoc:fix/paged-optimizer-non-cuda
9m 16s