AI/text-generation-webui
1
0
Fork 0
mirror of https://github.com/oobabooga/text-generation-webui.git synced 2024-10-01 01:26:03 -04:00
Code Issues Packages Projects Releases Wiki Activity

Improve readability

Browse Source
This commit is contained in:
oobabooga 2023-03-05 10:21:15 -03:00
parent 8e706df20e
commit a54b91af77
1 changed files with 0 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
modules/text_generation.py
View File

@ -115,7 +115,6 @@ def generate_reply(question, max_new_tokens, do_sample, temperature, top_p, typi
input_ids = encode(question, max_new_tokens) input_ids = encode(question, max_new_tokens)
cuda = "" if (shared.args.cpu or shared.args.deepspeed or shared.args.flexgen) else ".cuda()" cuda = "" if (shared.args.cpu or shared.args.deepspeed or shared.args.flexgen) else ".cuda()"
n = shared.tokenizer.eos_token_id if eos_token is None else encode(eos_token)[0][-1] n = shared.tokenizer.eos_token_id if eos_token is None else encode(eos_token)[0][-1]
if stopping_string is not None: if stopping_string is not None:
# The stopping_criteria code below was copied from # The stopping_criteria code below was copied from
# https://github.com/PygmalionAI/gradio-ui/blob/master/src/model.py # https://github.com/PygmalionAI/gradio-ui/blob/master/src/model.py
@ -152,14 +151,12 @@ def generate_reply(question, max_new_tokens, do_sample, temperature, top_p, typi
f"temperature={temperature}", f"temperature={temperature}",
f"stop={n}", f"stop={n}",
] ]
if shared.args.deepspeed: if shared.args.deepspeed:
generate_params.append("synced_gpus=True") generate_params.append("synced_gpus=True")
if shared.args.no_stream: if shared.args.no_stream:
generate_params.append("max_new_tokens=max_new_tokens") generate_params.append("max_new_tokens=max_new_tokens")
else: else:
generate_params.append("max_new_tokens=8") generate_params.append("max_new_tokens=8")
if shared.soft_prompt: if shared.soft_prompt:
inputs_embeds, filler_input_ids = generate_softprompt_input_tensors(input_ids) inputs_embeds, filler_input_ids = generate_softprompt_input_tensors(input_ids)
generate_params.insert(0, "inputs_embeds=inputs_embeds") generate_params.insert(0, "inputs_embeds=inputs_embeds")