OSError: [WinError -1073741795] Windows Error 0xc000001d And AttributeError: 'Llama' object has no attribute 'ctx'

[immmor]

D:\CursorFile\Python\privateGPT-main>python ingest.py Loading documents from source_documents Loaded 2 documents from source_documents Split into 603 chunks of text (max. 500 tokens each) llama.cpp: loading model from D:\CursorFile\Python\privateGPT-main\models\ggml-model-q4_0.bin Traceback (most recent call last): File "D:\anaconda3\lib\site-packages\langchain\embeddings\llamacpp.py", line 78, in validate_environment values["client"] = Llama( File "D:\anaconda3\lib\site-packages\llama_cpp\llama.py", line 155, in init self.ctx = llama_cpp.llama_init_from_file( File "D:\anaconda3\lib\site-packages\llama_cpp\llama_cpp.py", line 182, in llama_init_from_file return _lib.llama_init_from_file(path_model, params) OSError: [WinError -1073741795] Windows Error 0xc000001d

During handling of the above exception, another exception occurred:

Traceback (most recent call last): File "D:\CursorFile\Python\privateGPT-main\ingest.py", line 62, in main() File "D:\CursorFile\Python\privateGPT-main\ingest.py", line 53, in main llama = LlamaCppEmbeddings(model_path=llama_embeddings_model, n_ctx=model_n_ctx) File "pydantic\main.py", line 339, in pydantic.main.BaseModel.init File "pydantic\main.py", line 1102, in pydantic.main.validate_model File "D:\anaconda3\lib\site-packages\langchain\embeddings\llamacpp.py", line 98, in validate_environment raise NameError(f"Could not load Llama model from path: {model_path}") NameError: Could not load Llama model from path: D:\CursorFile\Python\privateGPT-main\models\ggml-model-q4_0.bin Exception ignored in: <function Llama.del at 0x0000017F4795CAF0> Traceback (most recent call last): File "D:\anaconda3\lib\site-packages\llama_cpp\llama.py", line 978, in del if self.ctx is not None: AttributeError: 'Llama' object has no attribute 'ctx'

[immmor]

When I run python ingest.py, I got these two error...

[nikhilsingh291]

The script tries to load models from the "models" directory in the code folder. You just need to download the model and put it in the directory. https://gpt4all.io/models/ggml-gpt4all-j-v1.3-groovy.bin https://huggingface.co/Pi3141/alpaca-native-7B-ggml/resolve/397e872bf4c83f4c642317a5bf65ce84a105786e/ggml-model-q4_0.bin

Above are the model links to download, present in the readme of the project.

[redfort1987]

I've done this, and its still giving me the error.

[redfort1987]

The models work with GPT4all, so it must be the privateGPT code.

[mrharrison007]

Getting "OSError: [WinError -1073741795] Windows Error 0xc000001d" also. My next step was to validate the LLMs. I'll try GPT4all to validate my downloaded LLMs.

[immmor]

The script tries to load models from the "models" directory in the code folder. You just need to download the model and put it in the directory. https://gpt4all.io/models/ggml-gpt4all-j-v1.3-groovy.bin https://huggingface.co/Pi3141/alpaca-native-7B-ggml/resolve/397e872bf4c83f4c642317a5bf65ce84a105786e/ggml-model-q4_0.bin

Above are the model links to download, present in the readme of the project.

I have downloaded those models and put them in 'models' folder. This error still happened.

[planerboy]

Have you found a solution? I am having the same problem

[LaszloA]

Same problem when running privateGPT.py : .... File ".../AppData\Local\Programs\Python\Python311\Lib\site-packages\gpt4all\pyllmodel.py", line 141, in load_model llmodel.llmodel_loadModel(self.model, model_path.encode('utf-8')) OSError: [WinError -1073741795] Windows Error 0xc000001d

Running Python 3.11 under Windows 10 x64, everything installed with default settings conforming to the tutorial Any idea for a fix?

[LaszloA]

Getting "OSError: [WinError -1073741795] Windows Error 0xc000001d" also. My next step was to validate the LLMs. I'll try GPT4all to validate my downloaded LLMs.

What do you mean by "validate the LLMs"?

[bones0]

I also have those windows errors with the version of gpt4all which does not cause the verification errors right away. So I assume this is the version which should work.

.conda\envs\gpt\lib\site-packages\gpt4all\pyllmodel.py", line 141, in load_model llmodel.llmodel_loadModel(self.model, model_path.encode('utf-8')) OSError: [WinError -1073741795] Windows Error 0xc000001d

Trying sfc /scannow right now. You never know...

edit: nope

[bones0]

https://stackoverflow.com/questions/53057591/importerror-dll-load-failed-with-error-code-1073741795 talks about old processors and downgrading tensorflow. We don't have tensorflow here, but still... I have a 10-year-old xeon. Not sure what to try out to fix it, though.

[bones0]

I installed gpt4all and the model downloader there issued several warnings that the bigger models need more RAM than I have. I am just guessing here - but could some windows errors occur because the model is simply using up all the RAM?

EDIT: The groovy-model is not maxing out the RAM. So it is not likely to be the problem here.

[bones0]

Interestingly, gpt4all also only reports "model loading error". No matter which one I try. I suspect there is something terminally messed up with Windows and I guess I will give up now. GPT apparently simply does not run here.

Edit: Yes, the "here" is the point. It works on the Core i7-5500, but not on the Xeon W3690. You might want to try different CPUs, just in case. It does seem to matter.