torch_dtype -> dtype (#3177)
* torch_dtype -> dtype * torch_dtype -> dtype
This commit is contained in:
@@ -85,9 +85,7 @@ def do_cli(model: Union[Path, str], output: Union[Path, str]) -> None:
|
|||||||
unpatch_llama4 = patch_llama4_linearized_modeling()
|
unpatch_llama4 = patch_llama4_linearized_modeling()
|
||||||
from transformers import Llama4ForConditionalGeneration
|
from transformers import Llama4ForConditionalGeneration
|
||||||
|
|
||||||
model_ = Llama4ForConditionalGeneration.from_pretrained(
|
model_ = Llama4ForConditionalGeneration.from_pretrained(model, dtype=torch.bfloat16)
|
||||||
model, torch_dtype=torch.bfloat16
|
|
||||||
)
|
|
||||||
processor = AutoProcessor.from_pretrained(model)
|
processor = AutoProcessor.from_pretrained(model)
|
||||||
processor.save_pretrained(output)
|
processor.save_pretrained(output)
|
||||||
|
|
||||||
|
|||||||
@@ -69,7 +69,7 @@ def do_quantize(
|
|||||||
config = AutoConfig.from_pretrained(model_path)
|
config = AutoConfig.from_pretrained(model_path)
|
||||||
torch_dtype = config.torch_dtype if hasattr(config, "torch_dtype") else None
|
torch_dtype = config.torch_dtype if hasattr(config, "torch_dtype") else None
|
||||||
model = AutoModelForCausalLM.from_pretrained(
|
model = AutoModelForCausalLM.from_pretrained(
|
||||||
model_path, device_map="auto", torch_dtype=torch_dtype
|
model_path, device_map="auto", dtype=torch_dtype
|
||||||
)
|
)
|
||||||
|
|
||||||
LOG.info(
|
LOG.info(
|
||||||
|
|||||||
@@ -148,7 +148,7 @@ def load_sharded_model(
|
|||||||
model = AutoModelForCausalLM.from_pretrained(
|
model = AutoModelForCausalLM.from_pretrained(
|
||||||
model_name,
|
model_name,
|
||||||
use_cache=False,
|
use_cache=False,
|
||||||
torch_dtype=torch.float32,
|
dtype=torch.float32,
|
||||||
_attn_implementation=model_config._attn_implementation,
|
_attn_implementation=model_config._attn_implementation,
|
||||||
trust_remote_code=cfg.trust_remote_code,
|
trust_remote_code=cfg.trust_remote_code,
|
||||||
)
|
)
|
||||||
@@ -158,7 +158,7 @@ def load_sharded_model(
|
|||||||
with init_empty_weights():
|
with init_empty_weights():
|
||||||
model = AutoModelForCausalLM.from_config(
|
model = AutoModelForCausalLM.from_config(
|
||||||
model_config,
|
model_config,
|
||||||
torch_dtype=torch_dtype,
|
dtype=torch_dtype,
|
||||||
trust_remote_code=cfg.trust_remote_code,
|
trust_remote_code=cfg.trust_remote_code,
|
||||||
)
|
)
|
||||||
return model
|
return model
|
||||||
|
|||||||
@@ -160,7 +160,7 @@ def test_geglu_model_integration():
|
|||||||
"""Test GeGLU activation with Gemma model."""
|
"""Test GeGLU activation with Gemma model."""
|
||||||
model = AutoModelForCausalLM.from_pretrained(
|
model = AutoModelForCausalLM.from_pretrained(
|
||||||
"trl-internal-testing/tiny-Gemma2ForCausalLM",
|
"trl-internal-testing/tiny-Gemma2ForCausalLM",
|
||||||
torch_dtype=torch.float16,
|
dtype=torch.float16,
|
||||||
device_map="cuda:0",
|
device_map="cuda:0",
|
||||||
)
|
)
|
||||||
peft_config = get_peft_config(
|
peft_config = get_peft_config(
|
||||||
|
|||||||
@@ -39,7 +39,7 @@ def model():
|
|||||||
dummy_model = AutoModelForCausalLM.from_pretrained(
|
dummy_model = AutoModelForCausalLM.from_pretrained(
|
||||||
"Qwen/Qwen2-0.5B",
|
"Qwen/Qwen2-0.5B",
|
||||||
device_map="auto",
|
device_map="auto",
|
||||||
torch_dtype=torch.bfloat16,
|
dtype=torch.bfloat16,
|
||||||
)
|
)
|
||||||
with torch.device(dummy_model.device):
|
with torch.device(dummy_model.device):
|
||||||
dummy_model.model.embed_tokens = torch.nn.Embedding(
|
dummy_model.model.embed_tokens = torch.nn.Embedding(
|
||||||
|
|||||||
Reference in New Issue
Block a user