Mirrors
/
hf_text-generation-inference
mirror of https://github.com/huggingface/text-generation-inference.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

fix dbrx & opt model prefix bug (#2201) 

* Update idefics_causal_lm.py

Fix syntax issues

* fix dbrx & opt model prefix bug
This commit is contained in:
icyboy™ 2024-07-08 15:01:14 +08:00 committed by GitHub
parent 05c094fcfa
commit 521d0d990f
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
2 changed files with 2 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
server/text_generation_server/models/custom_modeling/flash_dbrx_modeling.py
View File

@ -711,7 +711,7 @@ class FlashDbrxForCausalLM(torch.nn.Module):
else: else:
prefix = f"{prefix}.transformer" prefix = f"{prefix}.transformer"
self.model = DbrxModel(config, weights) self.model = DbrxModel(prefix, config, weights)
self.lm_head = SpeculativeHead.load( self.lm_head = SpeculativeHead.load(
config, config,
prefix="lm_head", prefix="lm_head",

2
server/text_generation_server/models/custom_modeling/opt_modeling.py
View File

@ -757,7 +757,7 @@ class OPTForCausalLM(OPTPreTrainedModel):
else: else:
prefix = f"{prefix}.model" prefix = f"{prefix}.model"
self.model = OPTModel(config, weights) self.model = OPTModel(prefix, config, weights)
self.lm_head = SpeculativeHead.load( self.lm_head = SpeculativeHead.load(
config, prefix=f"{prefix}.decoder.embed_tokens", weights=weights config, prefix=f"{prefix}.decoder.embed_tokens", weights=weights