issues
search
huggingface
/
optimum-quanto
A pytorch quantization backend for optimum
Apache License 2.0
833
stars
61
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Add repr for QuantizedTransformersModel
#357
imba-tjd
closed
13 hours ago
0
Clarification on Per-Channel vs. Per-Tensor Quantization for Weights and Activations
#356
kirkdort44
opened
1 day ago
0
QuantizedModelForCausalLM lost repr, cannot see model structure
#355
imba-tjd
closed
13 hours ago
1
reload quantized model after saving
#354
lucienfostier
opened
5 days ago
1
Is it GPU Compatability Issue?
#353
kamrul-NSL
opened
1 week ago
0
task not support: image-text-to-text.
#352
fangyangci
opened
1 week ago
0
Its not an issue! Quantization Granularity Question
#351
ClaraLovesFunk
opened
1 week ago
0
enable qbitstensor test on xpu
#350
dacorvo
closed
1 week ago
0
[tests] enable testing for xpu (rebased)
#349
dacorvo
closed
1 week ago
0
Quantized Flux pipeline to cuda?
#348
squewel
opened
2 weeks ago
1
Weights Still in FP32 after Quantization
#347
ClaraLovesFunk
opened
2 weeks ago
4
How to support activation 4bit quantization?
#346
Ther-nullptr
opened
3 weeks ago
1
[tests] enable `test_weight_qbits_tensor_linear_cuda` on xpu devices
#345
faaany
closed
1 week ago
1
[tests] enable testing for xpu
#344
faaany
closed
1 week ago
5
Only random noise is generated with Flux + LoRA with optimum-quanto >= 0.2.5
#343
nelapetrzelkova
opened
3 weeks ago
2
support
#342
werruww
closed
3 weeks ago
0
Support QLayerNorm without weights
#341
dacorvo
closed
3 weeks ago
0
Inference Speed Slowdown with Static Quantization
#340
ClaraLovesFunk
opened
3 weeks ago
2
Error when quantizing flux (already fixed in master)
#339
samedii
closed
3 weeks ago
1
fix: use reshape instead of view
#338
dacorvo
closed
1 month ago
0
when I use quanto lib to quantize the flux model, the activation layer's quantization on pipe.transformer failed
#337
DaaadShot
closed
1 month ago
1
Will module output not be quantized when the model is directly trained after Calibration?
#336
tusiqi1
closed
5 days ago
7
LayerNorm with None weight throws exception
#335
doctorpangloss
closed
3 weeks ago
2
Switched linters, black -> ruff
#334
ishandeva
closed
1 month ago
2
Add marlin int4 kernel
#333
dacorvo
closed
1 month ago
0
Corrupted outputs with Marlin int4 kernels as parallelization increases
#332
dacorvo
opened
1 month ago
4
optimum-quanto 0.25 requires ninja but 'pip check flux' reports 'ninja-1.11.1.1 is not supported on this platform'
#331
Davros666
closed
2 weeks ago
2
Add hip support
#330
dacorvo
closed
1 month ago
0
Refactor extensions
#329
dacorvo
closed
1 month ago
0
Remove overheads in library
#328
dacorvo
closed
1 month ago
0
issues with non-contiguous Tensor
#327
bghira
closed
1 month ago
1
Fix lumina
#326
dacorvo
closed
1 month ago
0
Fix missing call in QuantizedTransformersModel
#325
dacorvo
closed
1 month ago
0
Do the Linear layers quantized with W8A8 include output quantization layers?
#324
chenghuaWang
closed
1 month ago
9
refactor(library): reduce overhead in marlin op
#323
dacorvo
closed
1 month ago
0
mps low-bit kernels from torchao
#322
bghira
closed
3 weeks ago
2
Ci move
#321
glegendre01
closed
1 month ago
0
Stricter optimized tensor tests
#320
dacorvo
closed
2 months ago
0
Accuracy issue when using torch._int_mm on AMD CPUs
#319
dacorvo
opened
2 months ago
5
chore: minimal python version is 3.9
#318
dacorvo
closed
2 months ago
0
Refactor AWQ gemm
#317
dacorvo
closed
2 months ago
0
More refactoring
#316
dacorvo
closed
2 months ago
0
Add marlin int4 kernel
#315
dacorvo
closed
1 month ago
1
Refactor QBitsTensor subclasses
#314
dacorvo
closed
2 months ago
0
Does AWQ is officially supported now?
#313
lifelongeeek
closed
1 month ago
5
qint4 failed for diffusers: QBitsTensor cannot be changed
#312
liyihao1230
opened
2 months ago
2
quantize(model, weights=qint4, activations=qint8) produce weights with dtype = torch.uint8
#311
lifelongeeek
closed
2 months ago
2
feat: e4m3fnuz added
#310
dacorvo
closed
2 months ago
0
fix(library): disable int_mm for CPU
#309
dacorvo
closed
2 months ago
7
feat(examples): add image classification example using quantized vit …
#308
shovan777
closed
2 months ago
2
Next