vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

convert-new.py : minor fixes

Browse source
This commit is contained in:
Georgi Gerganov 2023-08-17 17:16:26 +03:00
parent 5d044403d3
commit 3839704062
No known key found for this signature in database
GPG key ID: 449E073F9DC10735
1 changed files with 3 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

6
convert-new.py
View file

@ -193,7 +193,7 @@ class Params:
n_ctx = 2048 if config["norm_eps"] == 1e-06 else 4096 # hack to determine LLaMA v1 vs v2 n_ctx = 2048 if config["norm_eps"] == 1e-06 else 4096 # hack to determine LLaMA v1 vs v2
n_ff = -1; n_ff = -1;
n_head = config["n_heads"]; n_head = config["n_heads"];
n_head_kv = config["n_kv_head"] if "n_kv_head" in config else n_head; n_head_kv = config["n_kv_heads"] if "n_kv_heads" in config else n_head;
f_norm_eps = config["norm_eps"]; f_norm_eps = config["norm_eps"];
if n_vocab == -1: if n_vocab == -1:
@ -796,9 +796,9 @@ class OutputFile:
def pick_output_type(model: LazyModel, output_type_str: Optional[str]) -> GGMLFileType: def pick_output_type(model: LazyModel, output_type_str: Optional[str]) -> GGMLFileType:
wq_type = model[NAMES[gguf.MODEL_TENSOR.ATTN_Q].format(bid=0)+".weight"].data_type wq_type = model[NAMES[gguf.MODEL_TENSOR.ATTN_Q].format(bid=0)+".weight"].data_type
if output_type_str == "f32" or (output_type_str is None and wq_type in (DT_F32, DT_BF16)): if output_type_str == "f32" or (output_type_str is None and wq_type == DT_F32):
return GGMLFileType.AllF32 return GGMLFileType.AllF32
if output_type_str == "f16" or (output_type_str is None and wq_type == DT_F16): if output_type_str == "f16" or (output_type_str is None and wq_type in (DT_F16, DT_BF16)):
return GGMLFileType.MostlyF16 return GGMLFileType.MostlyF16
name_to_type = {name: lazy_tensor.data_type for (name, lazy_tensor) in model.items()} name_to_type = {name: lazy_tensor.data_type for (name, lazy_tensor) in model.items()}