A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 16, 22:00 UTC
[Bugfix] GPTBigCodeForCausalLM: Remove lm_head from supported_lora_modules. (#6326)
Signed-off-by: Thomas Parnell <tpa@zurich.ibm.com> Co-authored-by: Travis Johnson <tsjohnso@us.ibm.com>
Thomas Parnell committed
Jul 11, 2024 at 16:05 UTC
8a1415cf776b2b902f6429ecfc325877b57cbefe
1 file changed
+1
-1
vllm/model_executor/models/gpt_bigcode.py
+1
-1
@@ -235,7 +235,7 @@ class GPTBigCodeModel(nn.Module):
235
class GPTBigCodeForCausalLM(nn.Module, SupportsLoRA):
236
packed_modules_mapping = {"c_attn": ["c_attn"]}
237
238
- supported_lora_modules = ["c_fc", "c_proj", "wte", "lm_head", "c_attn"]
238
+ supported_lora_modules = ["c_fc", "c_proj", "wte", "c_attn"]
239
240
embedding_modules = {
241
"wte": "input_embeddings",