Skip to content
This repository has been archived by the owner on Oct 25, 2024. It is now read-only.

Adapt INC autoround changes #1669

Merged
merged 6 commits into from
Jul 25, 2024
Merged

Adapt INC autoround changes #1669

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
5725923
adapt INC autoround changes
changwangss Jul 17, 2024
dd24ee1
autoround accuracy change
changwangss Jul 17, 2024
d8802e2
to enable ci
changwangss Jul 17, 2024
6a32972
update commit
changwangss Jul 22, 2024
3f8595c
skip ut
changwangss Jul 25, 2024
1a379fc
Update unit-test-optimize.yml
changwangss Jul 25, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion examples/huggingface/pytorch/code-generation/quantization/requirements.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -11,5 +11,5 @@ tiktoken #code_gen
neural-compressor
intel_extension_for_pytorch==2.3.0
git+https://github.com/huggingface/optimum-intel.git@50d867c13b22c22eda451ddb67bddb8159670f85
auto-round==0.2
git+https://github.com/intel/auto-round.git@24b2e74070f2b4e6f26ff069ec75af74cf5b177c
git+https://github.com/bigcode-project/bigcode-evaluation-harness@094c7cc197d13a53c19303865e2056f1c7488ac1
2 changes: 1 addition & 1 deletion examples/huggingface/pytorch/text-generation/quantization/requirements_GPU.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -12,6 +12,6 @@ bitsandbytes #baichuan
transformers_stream_generator
tiktoken #qwen
einops #qwen
auto-round
git+https://github.com/intel/auto-round.git@24b2e74070f2b4e6f26ff069ec75af74cf5b177c
git+https://github.com/intel/neural-compressor.git
lm-eval==0.4.3
2 changes: 1 addition & 1 deletion examples/huggingface/pytorch/text-generation/quantization/requirements_cpu_woq.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -11,7 +11,7 @@ transformers_stream_generator
tiktoken #qwen
einops #qwen
git+https://github.com/intel/neural-speed.git
auto-round==0.2
git+https://github.com/intel/auto-round.git@24b2e74070f2b4e6f26ff069ec75af74cf5b177c
git+https://github.com/intel/neural-compressor.git
lm-eval==0.4.3
huggingface_hub
4 changes: 2 additions & 2 deletions intel_extension_for_transformers/transformers/llm/quantization/utils.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -658,7 +658,7 @@ def convert_to_quantized_model(model, config, device="cpu"):
lr=config.lr,
minmax_lr=config.minmax_lr,
seqlen=config.seq_len,
n_samples=config.n_samples,
nsamples=config.n_samples,
iters=config.iters,
scale_dtype=config.scale_dtype,
)
Expand All @@ -672,7 +672,7 @@ def convert_to_quantized_model(model, config, device="cpu"):
dataset_name="NeelNanda/pile-10k",
seed=42,
bs=config.batch_size,
n_samples=config.n_samples)
nsamples=config.n_samples)
run_fn = run_fn_for_autoround
run_args = (dataloader,)
model = prepare(model=model, quant_config=quant_config)
Expand Down
2 changes: 1 addition & 1 deletion tests/requirements.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,13 +1,13 @@
--extra-index-url https://download.pytorch.org/whl/cpu
accelerate
auto-gptq
auto-round==0.2
bitsandbytes
datasets==2.16.1
einops
evaluate
gguf
git+https://github.com/huggingface/optimum-intel.git@50d867c13b22c22eda451ddb67bddb8159670f85
git+https://github.com/intel/auto-round.git@24b2e74070f2b4e6f26ff069ec75af74cf5b177c
git+https://github.com/intel/neural-compressor.git
git+https://github.com/intel/neural-speed.git
intel-extension-for-pytorch==2.3.0
Expand Down
Loading