kyy/llm_trainer
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

update gradio, support multiple resp in api

Browse Source
This commit is contained in:
hiyouga
2023-11-01 23:02:16 +08:00
parent d8cf8cfdeb
commit 84af10cec9
10 changed files with 54 additions and 42 deletions
Download Patch File Download Diff File Expand all files Collapse all files

7
src/llmtuner/tuner/core/loader.py
View File

@@ -83,7 +83,7 @@ def load_model_and_tokenizer(
config = AutoConfig.from_pretrained(model_to_load, **config_kwargs)
# Fix tokenizer (for ChatGLM2)
# Fix tokenizer (for ChatGLM2 and ChatGLM3)
if getattr(config, "model_type", None) == "chatglm":
tokenizer._pad = MethodType(PreTrainedTokenizerBase._pad, tokenizer)
@@ -193,7 +193,7 @@ def load_model_and_tokenizer(
if isinstance(model, PreTrainedModel) and "GenerationMixin" not in str(model.generate.__func__):
model.generate = MethodType(PreTrainedModel.generate, model)
# Fix LM head (for ChatGLM2)
# Fix LM head (for ChatGLM2 and ChatGLM3)
if getattr(config, "model_type", None) == "chatglm":
setattr(model, "lm_head", model.transformer.output_layer)
@@ -212,8 +212,7 @@ def load_model_and_tokenizer(
# Prepare model with valuehead for RLHF
if stage == "rm" or stage == "ppo":
model = AutoModelForCausalLMWithValueHead.from_pretrained(model)
model._keys_to_ignore_on_save = None
model: "AutoModelForCausalLMWithValueHead" = AutoModelForCausalLMWithValueHead.from_pretrained(model)
reset_logging()
if stage == "rm" and model_args.checkpoint_dir is not None: # load valuehead weights to evaluate reward model
logger.warning("Only the last checkpoint containing valuehead will be loaded.")