mirror of
https://github.com/oobabooga/text-generation-webui.git
synced 2024-11-25 17:29:22 +01:00
commit
e813b322cf
@ -577,7 +577,7 @@ def find_all_histories_with_first_prompts(state):
|
|||||||
data = json.load(f)
|
data = json.load(f)
|
||||||
|
|
||||||
first_prompt = ""
|
first_prompt = ""
|
||||||
if 'visible' in data and len(data['visible']) > 0:
|
if data and 'visible' in data and len(data['visible']) > 0:
|
||||||
if data['internal'][0][0] == '<|BEGIN-VISIBLE-CHAT|>':
|
if data['internal'][0][0] == '<|BEGIN-VISIBLE-CHAT|>':
|
||||||
if len(data['visible']) > 1:
|
if len(data['visible']) > 1:
|
||||||
first_prompt = html.unescape(data['visible'][1][0])
|
first_prompt = html.unescape(data['visible'][1][0])
|
||||||
|
@ -100,9 +100,11 @@ def eval_with_progress(self, tokens: Sequence[int]):
|
|||||||
|
|
||||||
|
|
||||||
def monkey_patch_llama_cpp_python(lib):
|
def monkey_patch_llama_cpp_python(lib):
|
||||||
|
if getattr(lib.Llama, '_is_patched', False):
|
||||||
|
# If the patch is already applied, do nothing
|
||||||
|
return
|
||||||
|
|
||||||
def my_generate(self, *args, **kwargs):
|
def my_generate(self, *args, **kwargs):
|
||||||
|
|
||||||
if shared.args.streaming_llm:
|
if shared.args.streaming_llm:
|
||||||
new_sequence = args[0]
|
new_sequence = args[0]
|
||||||
past_sequence = self._input_ids
|
past_sequence = self._input_ids
|
||||||
@ -116,3 +118,6 @@ def monkey_patch_llama_cpp_python(lib):
|
|||||||
lib.Llama.eval = eval_with_progress
|
lib.Llama.eval = eval_with_progress
|
||||||
lib.Llama.original_generate = lib.Llama.generate
|
lib.Llama.original_generate = lib.Llama.generate
|
||||||
lib.Llama.generate = my_generate
|
lib.Llama.generate = my_generate
|
||||||
|
|
||||||
|
# Set the flag to indicate that the patch has been applied
|
||||||
|
lib.Llama._is_patched = True
|
||||||
|
Loading…
Reference in New Issue
Block a user