Description
Describe the bug/ 问题描述 (Mandatory / 必填)
A clear and concise description of what the bug is.
在调用这个函数：
def get_code_completion(prompt: str, model, tokenizer, temperature: float) -> str:
"""Generate code completion for a given prompt"""
try:
model.eval()
input_ids=tokenizer(prompt, return_tensors="ms")#改成符合mindspore形式的张量
outputs = model.generate(
input_ids=tokenized_input["input_ids"]
max_new_tokens=MAX_NEW_TOKENS,
temperature=temperature,
top_k=TOP_K,
top_p=TOP_P,
do_sample=True,
no_repeat_ngram_size=NO_REPEAT_NGRAM_SIZE,
repetition_penalty=REPETITION_PENALTY,
)
ms.ms_memory_recycle()
return tokenizer.batch_decode(outputs, skip_special_tokens=False)[0]
except Exception as e:
print(f"Error during code generation: {str(e)}")#转化成字符串打印错误信息
raise
接收到下图中来自transformers的models里面llama模型的报错，显示split没有dim这一个参数。
-
Hardware Environment(
Ascend/GPU/CPU) / 硬件环境:
Ascend -
Software Environment / 软件环境 (Mandatory / 必填):
在华为modelarts和openI中均出现该错误。
-- MindSpore version (e.g., 1.7.0.Bxxx) :2.4.0
-- Python version (e.g., Python 3.7.5) :3.9
-- 镜像：mindspore_2.2.0-cann_7.0.1-py_3.9-euler_2.10.7-aarch64-snt9b -
Excute Mode / 执行模式 (Mandatory / 必填)(
PyNative/Graph):Graph
Expected behavior / 预期结果 (Mandatory / 必填)
A clear and concise description of what you expected to happen.
问答应当返回调用llama微调模型返回tokenizer.batch_decode(outputs, skip_special_tokens=False)[0]的结果。
Screenshots/ 日志 / 截图 (Mandatory / 必填)
If applicable, add screenshots to help explain your problem.
