CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-11-03 09:22:01 +00:00
Code Issues Packages Projects Releases Wiki Activity

convert-gptneox-h5-to-gguf.py : add file_type key

Browse Source
This commit is contained in:
klosax
2023-08-13 13:54:10 +02:00
committed by GitHub
parent bf2dad3100
commit 2827b840e4
1 changed files with 2 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
convert-gptneox-h5-to-gguf.py
View File

@@ -90,8 +90,9 @@ print("gguf: get model metadata")
llm_arch = "gptneox" llm_arch = "gptneox"
block_count = hparams["num_hidden_layers"] block_count = hparams["num_hidden_layers"]
gguf_writer.add_name(last_dir)
gguf_writer.add_architecture(llm_arch) gguf_writer.add_architecture(llm_arch)
gguf_writer.add_name(last_dir)
gguf_writer.add_file_type( "All tensors F32" if ftype == 0 else "Most tensors F16, some F32")
gguf_writer.add_context_length(llm_arch, hparams["max_position_embeddings"]) gguf_writer.add_context_length(llm_arch, hparams["max_position_embeddings"])
gguf_writer.add_embedding_length(llm_arch, hparams["hidden_size"]) gguf_writer.add_embedding_length(llm_arch, hparams["hidden_size"])
gguf_writer.add_block_count(llm_arch, block_count) gguf_writer.add_block_count(llm_arch, block_count)