token probs for non HF loaders (#3957)

This commit is contained in:
saltacc 2023-09-17 13:42:32 +00:00 committed by GitHub
parent 0668f4e67f
commit cd08eb0753
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
5 changed files with 53 additions and 5 deletions

View File

@ -208,3 +208,8 @@ class ExllamaModel:
ids = ids.view(1, -1)
return self.tokenizer.decode(ids)[0]
def get_logits(self, token_ids, **kwargs):
self.cache.current_seq_len = 0
self.model.forward(token_ids[:, :-1], self.cache, input_mask=None, preprocess_only=True)
return self.model.forward(token_ids[:, -1:], self.cache, **kwargs).float().cpu()

View File

@ -113,3 +113,8 @@ class Exllamav2Model:
ids = ids.view(1, -1)
return self.tokenizer.decode(ids)[0]
def get_logits(self, token_ids, **kwargs):
self.cache.current_seq_len = 0
self.model.forward(token_ids[:, :-1], self.cache, input_mask=None, preprocess_only=True)
return self.model.forward(token_ids[:, -1:], self.cache, input_mask=None, **kwargs).float().cpu()

View File

@ -1,6 +1,7 @@
import re
from functools import partial
import numpy as np
import torch
from modules import RoPE, shared
@ -100,6 +101,12 @@ class LlamaCppModel:
def decode(self, tokens):
return self.model.detokenize(tokens)
def get_logits(self, tokens):
self.model.eval(tokens)
logits = self.model._scores
logits = np.expand_dims(logits, 0) # batch dim is expected
return torch.tensor(logits, dtype=torch.float32)
def generate(self, prompt, state, callback=None):
LogitsProcessorList = llama_cpp_lib().LogitsProcessorList

View File

@ -1,13 +1,32 @@
import torch
from modules import sampler_hijack, shared
from modules.exllama import ExllamaModel
from modules.exllamav2 import Exllamav2Model
from modules.llamacpp_model import LlamaCppModel
from modules.logging_colors import logger
from modules.text_generation import generate_reply
global_scores = None
def get_next_logits(prompt, state, use_samplers, previous):
if shared.model is None:
logger.error("No model is loaded! Select one in the Model tab.")
return 'Error: No model is loaded1 Select one in the Model tab.', previous
is_non_hf_exllamav2 = isinstance(shared.model, Exllamav2Model)
is_non_hf_exllamav1 = isinstance(shared.model, ExllamaModel)
is_non_hf_llamacpp = isinstance(shared.model, LlamaCppModel)
if use_samplers:
if any([is_non_hf_exllamav2, is_non_hf_exllamav1, is_non_hf_llamacpp]):
logger.error("Sampler hijacking is not supported non-Huggingface loaders.")
# sampling is all done in c for exllama, so it is really hard to hijack
# it should be possible to hijack llamacpp sampler by hijacking all their sampling methods,
# but it is not implemented yet
return 'Error: Sampler hijacking is not supported non-Huggingface loaders. Please disable the "Use samplers" option.', previous
state['max_new_tokens'] = 1
state['auto_max_new_tokens'] = False
for _ in generate_reply(prompt, state):
@ -15,17 +34,29 @@ def get_next_logits(prompt, state, use_samplers, previous):
scores = sampler_hijack.global_scores[-1]
else:
tokens = shared.tokenizer.encode(prompt, return_tensors='pt').cuda()
output = shared.model(input_ids=tokens)
scores = output['logits'][-1][-1]
if is_non_hf_exllamav2 or is_non_hf_exllamav1:
tokens = shared.tokenizer.encode(prompt).cuda()
scores = shared.model.get_logits(tokens)[-1][-1]
elif is_non_hf_llamacpp:
tokens = shared.tokenizer.encode(prompt)
scores = shared.model.get_logits(tokens)[-1][-1]
else:
tokens = shared.tokenizer.encode(prompt, return_tensors='pt').cuda()
output = shared.model(input_ids=tokens)
scores = output['logits'][-1][-1]
probs = torch.softmax(scores, dim=-1, dtype=torch.float)
topk_values, topk_indices = torch.topk(probs, k=25, largest=True, sorted=True)
topk_values = [f"{float(i):.5f}" for i in topk_values]
if is_non_hf_exllamav1 or is_non_hf_llamacpp:
topk_indices = [i.expand((1, 1)) for i in topk_indices]
tokens = [shared.tokenizer.decode(i) for i in topk_indices]
if is_non_hf_llamacpp:
tokens = [i.decode('utf-8') for i in tokens] # llamacpp returns bytes, not str
output = ''
for row in list(zip(topk_values, tokens)):
output += f"{row[0]} - {row[1]}\n"
output += f"{row[0]} - {repr(row[1])[1:-1]}\n"
return output, previous

View File

@ -150,7 +150,7 @@ def get_token_ids(prompt):
output = ''
for row in list(zip(tokens, decoded_tokens)):
output += f"{str(int(row[0])).ljust(5)} - {row[1]}\n"
output += f"{str(int(row[0])).ljust(5)} - {repr(row[1])[1:-1]}\n"
return output