SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

chore: fix typos in tests directory (#36785)

Browse Source 
* chore: fix typos in tests directory

* chore: fix typos in tests directory

* chore: fix typos in tests directory

* chore: fix typos in tests directory

* chore: fix typos in tests directory

* chore: fix typos in tests directory

* chore: fix typos in tests directory
This commit is contained in:
Afanti
2025-03-18 17:31:13 +08:00
committed by GitHub
parent 7f5077e536
commit 19b9d8ae13
12 changed files with 25 additions and 25 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
tests/quantization/bnb/README.md
View File

@@ -13,7 +13,7 @@ The following is the recipe on how to effectively debug `bitsandbytes` integrati
The following instructions are tested with 2 NVIDIA-Tesla T4 GPUs. To run successfully `bitsandbytes` you would need a 8-bit core tensor supported GPU. Note that Turing, Ampere or newer architectures - e.g. T4, RTX20s RTX30s, A40-A100, A6000 should be supported.
## Virutal envs
## Virtual envs
```bash
conda create --name int8-testing python==3.8
@@ -61,7 +61,7 @@ This happens when some Linear weights are set to the CPU when using `accelerate`
Use the latest version of `accelerate` with a command such as: `pip install -U accelerate` and the problem should be solved.
### `Parameter has no attribue .CB`
### `Parameter has no attribute .CB`
Same solution as above.
@@ -71,7 +71,7 @@ Run your script by pre-pending `CUDA_LAUNCH_BLOCKING=1` and you should observe a
### `CUDA illegal memory error: an illegal memory access at line...`:
Check the CUDA verisons with:
Check the CUDA versions with:
```bash
nvcc --version
```

6
tests/quantization/bnb/test_4bit.py
View File

@@ -179,7 +179,7 @@ class Bnb4BitTest(Base4bitTest):
def test_original_dtype(self):
r"""
A simple test to check if the model succesfully stores the original dtype
A simple test to check if the model successfully stores the original dtype
"""
self.assertTrue(hasattr(self.model_4bit.config, "_pre_quantization_dtype"))
self.assertFalse(hasattr(self.model_fp16.config, "_pre_quantization_dtype"))
@@ -496,8 +496,8 @@ class Pipeline4BitTest(Base4bitTest):
def test_pipeline(self):
r"""
The aim of this test is to verify that the mixed 4bit is compatible with `pipeline` from transformers. Since
we used pipline for inference speed benchmarking we want to make sure that this feature does not break anything
on pipline.
we used pipeline for inference speed benchmarking we want to make sure that this feature does not break anything
on pipeline.
"""
# self._clear_cuda_cache()
self.pipe = pipeline(

6
tests/quantization/bnb/test_mixed_int8.py
View File

@@ -213,7 +213,7 @@ class MixedInt8Test(BaseMixedInt8Test):
def test_original_dtype(self):
r"""
A simple test to check if the model succesfully stores the original dtype
A simple test to check if the model successfully stores the original dtype
"""
self.assertTrue(hasattr(self.model_8bit.config, "_pre_quantization_dtype"))
self.assertFalse(hasattr(self.model_fp16.config, "_pre_quantization_dtype"))
@@ -655,8 +655,8 @@ class MixedInt8TestPipeline(BaseMixedInt8Test):
def test_pipeline(self):
r"""
The aim of this test is to verify that the mixed int8 is compatible with `pipeline` from transformers. Since
we used pipline for inference speed benchmarking we want to make sure that this feature does not break anything
on pipline.
we used pipeline for inference speed benchmarking we want to make sure that this feature does not break anything
on pipeline.
"""
# self._clear_cuda_cache()
self.pipe = pipeline(

8
tests/quantization/gptq/test_gptq.py
View File

@@ -167,7 +167,7 @@ class GPTQTest(unittest.TestCase):
def test_original_dtype(self):
r"""
A simple test to check if the model succesfully stores the original dtype
A simple test to check if the model successfully stores the original dtype
"""
self.assertTrue(hasattr(self.quantized_model.config, "_pre_quantization_dtype"))
self.assertFalse(hasattr(self.model_fp16.config, "_pre_quantization_dtype"))
@@ -261,7 +261,7 @@ class GPTQTest(unittest.TestCase):
if self.device_map == "cpu":
quant_type = "ipex" if is_ipex_available() else "torch"
else:
# We expecte tritonv2 to be used here, because exllama backend doesn't support packing https://github.com/ModelCloud/GPTQModel/issues/1354
# We expect tritonv2 to be used here, because exllama backend doesn't support packing https://github.com/ModelCloud/GPTQModel/issues/1354
# TODO: Remove this once GPTQModel exllama kernels supports packing
quant_type = "tritonv2"
quantized_model_from_saved = AutoModelForCausalLM.from_pretrained(
@@ -433,7 +433,7 @@ class GPTQTestExllamaV2(unittest.TestCase):
"exllamav2",
)
else:
# We expecte tritonv2 to be used here, because exllama backend doesn't support packing https://github.com/ModelCloud/GPTQModel/issues/1354
# We expect tritonv2 to be used here, because exllama backend doesn't support packing https://github.com/ModelCloud/GPTQModel/issues/1354
# TODO: Remove this once GPTQModel exllama kernels supports packing
self.assertEqual(
self.quantized_model.model.layers[0].self_attn.k_proj.QUANT_TYPE,
@@ -458,7 +458,7 @@ class GPTQTestExllamaV2(unittest.TestCase):
def test_generate_quality(self):
"""
Simple test to check the quality of the model by comapring the the generated tokens with the expected tokens
Simple test to check the quality of the model by comparing the the generated tokens with the expected tokens
"""
self.check_inference_correctness(self.quantized_model)

2
tests/quantization/higgs/test_higgs.py
View File

@@ -184,7 +184,7 @@ class HiggsTest(unittest.TestCase):
output = model.generate(**input_ids, max_new_tokens=self.max_new_tokens)
self.assertEqual(self.tokenizer.decode(output[0], skip_special_tokens=True), self.EXPECTED_OUTPUT)
@unittest.skip("This will almost surely OOM. Enable when swithed to a smaller model")
@unittest.skip("This will almost surely OOM. Enable when switched to a smaller model")
def test_dequantize(self):
"""
Test the ability to dequantize a model

4
tests/quantization/torchao_integration/test_torchao.py
View File

@@ -202,7 +202,7 @@ class TorchAoGPUTest(TorchAoTest):
def test_int4wo_offload(self):
"""
Simple test that checks if the quantized model int4 wieght only is working properly with cpu/disk offload
Simple test that checks if the quantized model int4 weight only is working properly with cpu/disk offload
"""
device_map_offload = {
@@ -254,7 +254,7 @@ class TorchAoGPUTest(TorchAoTest):
@require_torch_multi_gpu
def test_int4wo_quant_multi_gpu(self):
"""
Simple test that checks if the quantized model int4 wieght only is working properly with multiple GPUs
Simple test that checks if the quantized model int4 weight only is working properly with multiple GPUs
set CUDA_VISIBLE_DEVICES=0,1 if you have more than 2 GPUS
"""