Skip to content

GitLab

Commit d8b9f285 authored by zhouxiang's avatar zhouxiang
Browse files

更新修复偶发的持续输出问题

parent a9bd16ac
Hide whitespace changes
Inline Side-by-side
Showing with 281 additions and 159 deletions
README.md
View file @ d8b9f285
......@@ -3,7 +3,7 @@
## 模型介绍
通义千问-7B（Qwen-7B） 是阿里云研发的通义千问大模型系列的70亿参数规模的模型。Qwen-7B是基于Transformer的大语言模型, 在超大规模的预训练数据上进行训练得到。预训练数据类型多样，覆盖广泛，包括大量网络文本、专业书籍、代码等。同时，在Qwen-7B的基础上，使用对齐机制打造了基于大语言模型的AI助手Qwen-7B-Chat。
本项目主要针对Qwen-7B-Chat在DCU平台的推理性能优化，达到DCU平台较快的对话效果。)
本项目主要针对Qwen-7B-Chat在DCU平台的推理性能优化，达到DCU平台较快的对话效果。
## 模型结构
......
cli_demo.py
View file @ d8b9f285
......@@ -24,7 +24,7 @@ if __name__ == "__main__":
continue
print("AI:", end = "")
curResponse = ""
for response in model.stream_response(query, history = history):
for response in model.stream_response(query, history = history, do_sample = True, top_p = 0.8, top_k = 1, temperature = 1.0, repeat_penalty = 1.0):
curResponse += response
print(response, flush = True, end = "")
history.append((query, curResponse))
\ No newline at end of file
package/fastllm_pytools/hf_model.py
View file @ d8b9f285
from fastllm_pytools import llm
import torch
import ctypes
import numpy as np
from fastllm_pytools import llm;
import torch;
import ctypes;
import numpy as np;
fastllm_data_type_dict = {
"int4": 8,
......@@ -22,60 +22,61 @@ def create(model,
history_sep = None,
dtype = "float16"):
if (dtype not in fastllm_data_type_dict):
print("dtype should in ", list(fastllm_data_type_dict.keys()))
exit(0)
print("dtype should in ", list(fastllm_data_type_dict.keys()));
exit(0);
# 0.1 model info
modelInfo = model.config.__dict__
if model.generation_config is not None:
modelInfo.update(model.generation_config.__dict__)
if (pre_prompt):
modelInfo["pre_prompt"] = pre_prompt
modelInfo["pre_prompt"] = pre_prompt;
if (user_role):
modelInfo["user_role"] = user_role
modelInfo["user_role"] = user_role;
if (bot_role):
modelInfo["bot_role"] = bot_role
modelInfo["bot_role"] = bot_role;
if (history_sep):
modelInfo["history_sep"] = history_sep
modelInfo["history_sep"] = history_sep;
if (modelInfo["model_type"] == "baichuan" and hasattr(model, "model") and hasattr(model.model, "get_alibi_mask")):
# Baichuan 2代
modelInfo["use_alibi"] = "1"
modelInfo["pre_prompt"] = ""
modelInfo["user_role"] = ("<FLM_FIX_TOKEN_" + str(model.generation_config.user_token_id) + "> ") if hasattr(model.generation_config, "user_token_id") else ""
modelInfo["bot_role"] = ("<FLM_FIX_TOKEN_" + str(model.generation_config.assistant_token_id) + ">") if hasattr(model.generation_config, "assistant_token_id") else ""
modelInfo["history_sep"] = ""
modelInfo["use_alibi"] = "1";
modelInfo["pre_prompt"] = "";
modelInfo["user_role"] = ("<FLM_FIX_TOKEN_" + str(model.generation_config.user_token_id) + "> ") if hasattr(model.generation_config, "user_token_id") else "";
modelInfo["bot_role"] = ("<FLM_FIX_TOKEN_" + str(model.generation_config.assistant_token_id) + ">") if hasattr(model.generation_config, "assistant_token_id") else "";
modelInfo["history_sep"] = "";
if (modelInfo["model_type"] == "qwen"):
if modelInfo["chat_format"] == "chatml":
modelInfo["im_end_id"] = tokenizer.im_end_id
modelInfo["im_start_id"] = tokenizer.im_start_id
weight_type_dict = {}
module_dict = {}
weight_bits = {}
weight_type_dict = {};
module_dict = {};
weight_bits = {};
for key, m in model.named_modules():
if (str(type(m)).find("QuantizedLinear") != -1):
weight_type_dict[key + ".weight"] = "QuantizedLinear"
weight_bits[key + ".weight"] = m.weight_bit_width
weight_type_dict[key + ".weight"] = "QuantizedLinear";
weight_bits[key + ".weight"] = m.weight_bit_width;
if (isinstance(m, torch.nn.Linear)):
weight_type_dict[key + ".weight"] = "linear"
module_dict[key + ".weight"] = m
weight_type_dict[key + ".weight"] = "linear";
module_dict[key + ".weight"] = m;
if (isinstance(m, torch.nn.Embedding)):
weight_type_dict[key] = "embedding"
weight_type_dict[key] = "embedding";
peft_config = {}
active_adapter = ""
if hasattr(model, "peft_config"):
peft_config = model.peft_config
if hasattr(model, "active_adapter"):
if hasattr(model, "active_adapter") and isinstance(model.active_adapter, str):
# in transformers >= 4.33.0, active_adapter is a funtion in model, ignore it now
active_adapter = model.active_adapter
model = model.cpu()
dict = model.state_dict()
model_type = model.config.__dict__["model_type"]
model = llm.fastllm_lib.create_empty_llm_model(model_type.encode())
model = model.cpu();
dict = model.state_dict();
model_type = model.config.__dict__["model_type"];
model = llm.fastllm_lib.create_empty_llm_model(model_type.encode());
for it in modelInfo.keys():
llm.fastllm_lib.add_dict_llm_model(model, str(it).encode(), str(modelInfo[it]).encode())
llm.fastllm_lib.add_dict_llm_model(model, str(it).encode(), str(modelInfo[it]).encode());
for adapter_name in peft_config.keys():
adapter_dict = peft_config[adapter_name].__dict__
......@@ -90,39 +91,39 @@ def create(model,
if modelInfo["model_type"] == "qwen":
pass
else:
tokenizer = tokenizer.tokenizer
tokenizer = tokenizer.tokenizer;
if (hasattr(tokenizer, "sp_model")):
piece_size = tokenizer.sp_model.piece_size()
piece_size = tokenizer.sp_model.piece_size();
for i in range(piece_size):
llm.fastllm_lib.add_tokenizer_word_llm_model(model, tokenizer.sp_model.id_to_piece(i).encode(),
i, ctypes.c_float(tokenizer.sp_model.get_score(i)))
i, ctypes.c_float(tokenizer.sp_model.get_score(i)));
else:
vocab = tokenizer.get_vocab()
vocab = tokenizer.get_vocab();
for v in vocab.keys():
if (modelInfo["model_type"] == "moss"):
vv = [(ord(c) if c not in tokenizer.byte_decoder else tokenizer.byte_decoder[c]) for c in v]
llm.fastllm_lib.add_tokenizer_word_llm_model(model, vv, vocab[v], ctypes.c_float(1.0))
vv = [(ord(c) if c not in tokenizer.byte_decoder else tokenizer.byte_decoder[c]) for c in v];
llm.fastllm_lib.add_tokenizer_word_llm_model(model, vv, vocab[v], ctypes.c_float(1.0));
elif (modelInfo["model_type"] == "qwen"):
llm.fastllm_lib.add_tokenizer_word_llm_model(model, v, vocab[v], ctypes.c_float(1.0))
llm.fastllm_lib.add_tokenizer_word_llm_model(model, v, vocab[v], ctypes.c_float(1.0));
else:
llm.fastllm_lib.add_tokenizer_word_llm_model(model, v.encode(), vocab[v], ctypes.c_float(1.0))
tot = 0
llm.fastllm_lib.add_tokenizer_word_llm_model(model, v.encode(), vocab[v], ctypes.c_float(1.0));
tot = 0;
for key in dict:
ori_data_type = 0
ori_np_data_type = np.float32
cur_weight_type = 0
ori_data_type = 0;
ori_np_data_type = np.float32;
cur_weight_type = 0;
if (key in weight_type_dict and weight_type_dict[key] in fastllm_weight_type_dict):
cur_weight_type = fastllm_weight_type_dict[weight_type_dict[key]]
to_data_type = 0
cur_weight_type = fastllm_weight_type_dict[weight_type_dict[key]];
to_data_type = 0;
if (cur_weight_type == 1):
to_data_type = fastllm_data_type_dict[dtype]
to_data_type = fastllm_data_type_dict[dtype];
if (to_data_type == 7):
ori_data_type = 7
ori_np_data_type = np.float16
ori_data_type = 7;
ori_np_data_type = np.float16;
elif (cur_weight_type == 2):
# TODO bfloat
to_data_type = 0
to_data_type = 0;
weight_name = key
if peft_config is not None:
......@@ -133,19 +134,19 @@ def create(model,
(ctypes.c_int * len(dict[key].shape))(*list(dict[key].shape)),
weight_bits[key],
dict[key + "_scale"].numpy().astype(np.float32).ctypes.data_as(ctypes.c_void_p),
dict[key].numpy().ctypes.data_as(ctypes.c_void_p))
dict[key].numpy().ctypes.data_as(ctypes.c_void_p));
else:
llm.fastllm_lib.add_weight_llm_model(model, weight_name.encode(),
len(dict[key].shape),
(ctypes.c_int * len(dict[key].shape))(*list(dict[key].shape)),
to_data_type, cur_weight_type, ori_data_type,
dict[key].numpy().astype(ori_np_data_type).ctypes.data_as(ctypes.c_void_p))
tot += 1
print("convert (", tot, "/", len(dict), end = " )\r")
dict[key].numpy().astype(ori_np_data_type).ctypes.data_as(ctypes.c_void_p));
tot += 1;
print("convert (", tot, "/", len(dict), end = " )\r");
print("")
llm.fastllm_lib.init_params_llm_model(model)
llm.fastllm_lib.warmup_llm_model(model)
ret = llm.model("", id = model)
return ret
print("");
llm.fastllm_lib.init_params_llm_model(model);
llm.fastllm_lib.warmup_llm_model(model);
ret = llm.model("", id = model);
return ret;
package/fastllm_pytools/llm.py
View file @ d8b9f285
import ctypes
import os
from typing import Optional, Tuple, Union, List, Callable, Dict, Any
import ctypes;
import math
import os;
import threading
from typing import Optional, Tuple, Union, List, Callable, Dict, Any;
import platform
if platform.system() == 'Windows':
......@@ -11,6 +13,12 @@ else:
fastllm_lib.create_llm_model.argtypes = [ctypes.c_char_p]
fastllm_lib.create_llm_model.restype = ctypes.c_int
fastllm_lib.token_decode.argtypes = [ctypes.c_int, ctypes.c_int, ctypes.c_int, ctypes.c_char_p]
fastllm_lib.token_decode.restype = ctypes.c_int
fastllm_lib.token_encode_string.argtypes = [ctypes.c_int, ctypes.c_char_p, ctypes.c_int, ctypes.POINTER(ctypes.c_int)]
fastllm_lib.token_encode_string.restype = ctypes.c_int
fastllm_lib.launch_response_llm_model.argtypes = [ctypes.c_int, ctypes.c_int, ctypes.c_void_p,
ctypes.c_int, ctypes.c_bool, ctypes.c_float, ctypes.c_int,
ctypes.c_float, ctypes.c_float, ctypes.c_bool]
......@@ -46,106 +54,178 @@ fastllm_lib.add_tokenizer_word_llm_model.argtype = [ctypes.c_int, ctypes.c_char_
fastllm_lib.set_device_map.argtype = [ctypes.c_int, ctypes.c_void_p, ctypes.c_char_p, ctypes.c_void_p]
def set_cpu_threads(threads: int):
fastllm_lib.set_cpu_threads(threads)
fastllm_lib.set_cpu_threads(threads);
def get_cpu_threads() -> int:
return fastllm_lib.get_cpu_threads()
return fastllm_lib.get_cpu_threads();
def print_ins_info():
fastllm_lib.print_cpu_ins()
fastllm_lib.print_cpu_ins();
def set_cpu_kvcache(cpu_kvcache):
fastllm_lib.set_kvcache_in_cpu(ctypes.c_bool(cpu_kvcache))
fastllm_lib.set_kvcache_in_cpu(ctypes.c_bool(cpu_kvcache));
def get_cpu_kvcache():
return fastllm_lib.get_kvcache_in_cpu()
return fastllm_lib.get_kvcache_in_cpu();
def set_cpu_low_mem(low_mem):
fastllm_lib.set_cpu_low_mem(ctypes.c_bool(low_mem))
fastllm_lib.set_cpu_low_mem(ctypes.c_bool(low_mem));
def get_cpu_low_mem():
return fastllm_lib.get_cpu_low_mem()
return fastllm_lib.get_cpu_low_mem();
def set_device_map(device_map):
devices = []
values = []
devices = [];
values = [];
if (isinstance(device_map, str)):
devices.append(device_map)
values.append(1)
devices.append(device_map);
values.append(1);
elif (isinstance(device_map, list)):
devices = [str(x) for x in device_map]
values = [1 for x in device_map]
devices = [str(x) for x in device_map];
values = [1 for x in device_map];
elif (isinstance(device_map, dict)):
devices = [str(x) for x in device_map.keys()]
values = [int(device_map[x]) for x in device_map.keys()]
devices = [str(x) for x in device_map.keys()];
values = [int(device_map[x]) for x in device_map.keys()];
else:
print("set_device_map error.")
return
device_str = ''.join(devices)
device_len = [len(x) for x in devices]
print("set_device_map error.");
return;
device_str = ''.join(devices);
device_len = [len(x) for x in devices];
fastllm_lib.set_device_map(len(device_len),
(ctypes.c_int * len(device_len))(*device_len),
device_str.encode(),
(ctypes.c_int * len(values))(*values))
(ctypes.c_int * len(values))(*values));
def from_hf(model,
tokenizer = None,
dtype = "float16"):
from fastllm_pytools import hf_model
return hf_model.create(model, tokenizer, dtype = dtype)
from fastllm_pytools import hf_model;
return hf_model.create(model, tokenizer, dtype = dtype);
class model:
def __init__ (self, path : str,
id : int = -99999):
if (id != -99999):
self.model = id
self.model = id;
else:
self.model = fastllm_lib.create_llm_model(path.encode())
self.direct_query = False
self.model = fastllm_lib.create_llm_model(path.encode());
self.direct_query = False;
# 为了减少重复申请释放buffer对象而使用的线程局部存储区对象池
self.thread_local_obj = threading.local()
self.thread_local_obj.tokenizer_encode_string__output_buffer = None
self.thread_local_obj.tokenizer_decode_token__output_buffer = None
# tokenizer_decode_token 输出结果的静态缓存，手工触发构建
# 由于token数量有限且不太多，所以缓存该结果来减少调用较为适合。
# 不做成自动缓存是为了避免在多线程调用的时候对缓存dict加锁，同时也为不同场景提供选择空间
self.tokenizer_decode_token_cache = None
def get_prompt(self,
query: str,
history: List[Tuple[str, str]] = None) -> str:
if (not(history)):
history = []
prompt = ""
history = [];
prompt = "";
for i, (old_query, response) in enumerate(history):
prompt = fastllm_lib.make_history_llm_model(self.model, prompt.encode(), i, old_query.encode(), response.encode()).decode()
prompt = fastllm_lib.make_input_llm_model(self.model, prompt.encode(), len(history), query.encode()).decode()
return prompt
prompt = fastllm_lib.make_history_llm_model(self.model, prompt.encode(), i, old_query.encode(), response.encode()).decode();
prompt = fastllm_lib.make_input_llm_model(self.model, prompt.encode(), len(history), query.encode()).decode();
return prompt;
def save(self, path : str):
fastllm_lib.save_llm_model(self.model, path.encode())
fastllm_lib.save_llm_model(self.model, path.encode());
def eval(self):
pass
pass;
def build_tokenizer_decode_token_cache(self):
if self.tokenizer_decode_token_cache is not None:
return
cache_dict = dict()
vocab_size = fastllm_lib.get_tokenizer_vocab_size(self.model)
for token_id in range(vocab_size):
cache_dict[token_id] = self.tokenizer_decode_token(token_id)
self.tokenizer_decode_token_cache = cache_dict
def tokenizer_encode_string(self, content: str) -> List[int]:
output_buffer_init_len = 1024
if self.thread_local_obj.tokenizer_encode_string__output_buffer is None:
self.thread_local_obj.tokenizer_encode_string__output_buffer = (ctypes.c_int * output_buffer_init_len)()
buffer = self.thread_local_obj.tokenizer_encode_string__output_buffer
buffer_len = len(buffer)
result_len = fastllm_lib.token_encode_string(self.model, content.encode(), buffer_len, buffer)
if result_len > buffer_len:
if result_len > 10240:
# 要处理的数据过长，使用一次性的buffer
temp_buffer = (ctypes.c_int * result_len)()
ret = fastllm_lib.token_encode_string(self.model, content.encode(), result_len, temp_buffer)
return [i for i in temp_buffer]
else:
# 扩展buffer大小
new_buffer_len = round(math.ceil(result_len / 1024.0)) * 1024
buffer = (ctypes.c_int * new_buffer_len)()
self.thread_local_obj.tokenizer_encode_string__output_buffer = buffer
result_len = fastllm_lib.token_encode_string(self.model, content.encode(), new_buffer_len, buffer)
return [buffer[i] for i in range(result_len)]
def tokenizer_decode_token(self, token_id: int) -> bytes:
if self.tokenizer_decode_token_cache is not None:
cache_result = self.tokenizer_decode_token_cache.get(token_id)
if cache_result is not None:
return cache_result
output_buffer_init_len = 256
if self.thread_local_obj.tokenizer_decode_token__output_buffer is None:
self.thread_local_obj.tokenizer_decode_token__output_buffer = ctypes.create_string_buffer(output_buffer_init_len)
buffer = self.thread_local_obj.tokenizer_decode_token__output_buffer
ret = fastllm_lib.token_decode(self.model, token_id, len(buffer), buffer)
if ret > 0:
# buffer长度不够，扩展buffer大小
new_buffer_len = round(math.ceil(ret / 16.0)) * 16
buffer = ctypes.create_string_buffer(new_buffer_len)
self.thread_local_obj.tokenizer_decode_token__output_buffer = buffer
ret = fastllm_lib.token_decode(self.model, token_id, len(buffer), buffer)
assert ret == 0
buffer_bytes = buffer.raw
result_len = len(buffer_bytes)
for i in range(len(buffer_bytes)):
if buffer_bytes[i] == 0:
result_len = i
break
return buffer_bytes[:result_len]
def response_logits(self,
query: str,
history: List[Tuple[str, str]] = None,
tokenizer = None) -> str:
prompt = query if self.direct_query else self.get_prompt(query, history)
prompt = query if self.direct_query else self.get_prompt(query, history);
if (tokenizer == None):
handle = fastllm_lib.launch_response_str_llm_model(self.model, prompt.encode(),
ctypes.c_int(1), ctypes.c_bool(False), ctypes.c_float(1), ctypes.c_int(1),
ctypes.c_float(1), ctypes.c_float(1), ctypes.c_bool(True))
ctypes.c_float(1), ctypes.c_float(1), ctypes.c_bool(True));
else:
input = tokenizer.encode(prompt)
input = tokenizer.encode(prompt);
handle = fastllm_lib.launch_response_llm_model(self.model, len(input), (ctypes.c_int * len(input))(*input),
1, False, 1, 1, 1, 1, True)
vocab_size = fastllm_lib.get_tokenizer_vocab_size(self.model)
1, False, 1, 1, 1, 1, True);
vocab_size = fastllm_lib.get_tokenizer_vocab_size(self.model);
logits = list(range(vocab_size))
array = (ctypes.c_float * (vocab_size * 4))(*logits)
ret = fastllm_lib.fetch_response_logits_llm_model(self.model, handle, array)
out = list(array)[:vocab_size]
array = (ctypes.c_float * (vocab_size * 4))(*logits);
ret = fastllm_lib.fetch_response_logits_llm_model(self.model, handle, array);
out = list(array)[:vocab_size];
while (ret != -1):
ret = fastllm_lib.fetch_response_logits_llm_model(self.model, handle, array)
return out
ret = fastllm_lib.fetch_response_logits_llm_model(self.model, handle, array);
return out;
def response(self,
query: str,
history: List[Tuple[str, str]] = None,
max_length: int = 8192, do_sample = True, top_p = 0.8, top_k = 1, temperature = 1.0, repeat_penalty = 1.0) -> str:
ret = ""
ret = "";
for i in self.stream_response(query = query,
history = history,
max_length = max_length,
......@@ -154,84 +234,111 @@ class model:
temperature = temperature,
repeat_penalty = repeat_penalty,
one_by_one = True):
ret += i
return ret
ret += i;
return ret;
def stream_response(self,
query: str,
history: List[Tuple[str, str]] = None,
max_length: int = 8192, do_sample = True, top_p = 0.8, top_k = 1, temperature = 1.0, repeat_penalty = 1.0,
one_by_one = True):
prompt = query if self.direct_query else self.get_prompt(query, history)
prompt = query if self.direct_query else self.get_prompt(query, history);
handle = fastllm_lib.launch_response_str_llm_model(self.model, prompt.encode(),
ctypes.c_int(max_length), ctypes.c_bool(do_sample), ctypes.c_float(top_p), ctypes.c_int(top_k),
ctypes.c_float(temperature), ctypes.c_float(repeat_penalty), ctypes.c_bool(False))
res = ""
ret = b''
fail_cnt = 0
ctypes.c_float(temperature), ctypes.c_float(repeat_penalty), ctypes.c_bool(False));
res = "";
ret = b'';
fail_cnt = 0;
while True:
ret += fastllm_lib.fetch_response_str_llm_model(self.model, handle)
cur = ""
ret += fastllm_lib.fetch_response_str_llm_model(self.model, handle);
cur = "";
try:
cur = ret.decode()
ret = b''
cur = ret.decode();
ret = b'';
except:
fail_cnt += 1
fail_cnt += 1;
if (fail_cnt == 20):
break
break;
else:
continue
fail_cnt = 0
continue;
fail_cnt = 0;
if (cur == "<flmeos>"):
break;
if one_by_one:
yield cur;
else:
res += cur;
yield res;
def stream_response_raw(self,
input_tokens: List[int],
max_length: int = 8192, do_sample = True, top_p = 0.8, top_k = 1, temperature = 1.0, repeat_penalty = 1.0,
one_by_one = True
):
handle = fastllm_lib.launch_response_llm_model(self.model, len(input_tokens),
(ctypes.c_int * len(input_tokens))(*input_tokens),
ctypes.c_int(max_length), ctypes.c_bool(do_sample), ctypes.c_float(top_p), ctypes.c_int(top_k),
ctypes.c_float(temperature), ctypes.c_float(repeat_penalty), ctypes.c_bool(False))
# 可能遇到长尾char需要多个token才能够生成，所以只返回bytes，string.decode策略交给外部
# 方便统计输出token数量，和控制不完整utf8时候解码的逻辑
total_bytes = b''
while True:
cur_token = fastllm_lib.fetch_response_llm_model(self.model, handle)
if cur_token == -1:
break
cur_bytes = self.tokenizer_decode_token(cur_token)
if one_by_one:
yield cur
yield cur_bytes
else:
res += cur
yield res
total_bytes += cur_bytes
yield total_bytes
def chat(self, tokenizer, query: str, history: List[Tuple[str, str]] = None, max_length: int = 8192,
do_sample = True, top_p = 0.8, top_k = 1, temperature = 1.0, repeat_penalty = 1.0, **kwargs):
if (not(history)):
history = []
prompt = query if self.direct_query else self.get_prompt(query, history)
input = tokenizer.encode(prompt)
history = [];
prompt = query if self.direct_query else self.get_prompt(query, history);
input = tokenizer.encode(prompt);
handle = fastllm_lib.launch_response_llm_model(self.model, len(input), (ctypes.c_int * len(input))(*input),
max_length, do_sample, top_p, top_k, temperature, repeat_penalty,
False)
False);
result = []
result = [];
while True:
cur = fastllm_lib.fetch_response_llm_model(self.model, handle)
cur = fastllm_lib.fetch_response_llm_model(self.model, handle);
if (cur == -1):
break
result.append(cur)
response = tokenizer.decode(result)
history = history + [(query, response)]
return response, history
break;
result.append(cur);
response = tokenizer.decode(result);
history = history + [(query, response)];
return response, history;
def stream_chat(self, tokenizer, query: str, history: List[Tuple[str, str]] = None, past_key_values = None,
max_length: int = 8192, do_sample = True, top_p = 0.8, top_k = 1, temperature = 1.0, repeat_penalty = 1.0,
return_past_key_values = False, **kwargs) -> str:
if (not(history)):
history = []
prompt = query if self.direct_query else self.get_prompt(query, history)
input = tokenizer.encode(prompt)
history = [];
prompt = query if self.direct_query else self.get_prompt(query, history);
input = tokenizer.encode(prompt);
handle = fastllm_lib.launch_response_llm_model(self.model, len(input), (ctypes.c_int * len(input))(*input),
max_length, do_sample, top_p, top_k, temperature, repeat_penalty,
False)
tokens = []
False);
tokens = [];
while True:
cur = fastllm_lib.fetch_response_llm_model(self.model, handle)
cur = fastllm_lib.fetch_response_llm_model(self.model, handle);
if (cur == -1):
break
tokens.append(cur)
response = tokenizer.decode(tokens)
new_history = history + [(query, response)]
break;
tokens.append(cur);
response = tokenizer.decode(tokens);
new_history = history + [(query, response)];
if return_past_key_values:
yield response, new_history, None
yield response, new_history, None;
else:
yield response, new_history
yield response, new_history;
def set_adapter(self, name: str):
fastllm_lib.set_adapter(self.model, str(name).encode())
......
package/fastllm_pytools/torch2flm.py
View file @ d8b9f285
......@@ -21,8 +21,8 @@ fastllm_weight_type_dict = {
"embedding": 2
}
v = np.random.randint(-127, 127, [10, 20])
temp = v
v = np.random.randint(-127, 127, [10, 20]);
temp = v;
c_max = np.expand_dims(np.abs(v).max(axis = -1), -1)
c_scale = c_max / 127.0
v = (v / c_scale + 128.5).clip(1, 255).astype(np.uint8)
......@@ -34,23 +34,31 @@ def write_int8(fo, v):
fo.write(struct.pack('i', 3))
fo.write(struct.pack('i', 0))
for i in range(c_max.shape[0]):
fo.write(struct.pack('f', -c_max[i][0]))
fo.write(struct.pack('f', c_max[i][0]))
fo.write(struct.pack('f', -c_max[i][0]));
fo.write(struct.pack('f', c_max[i][0]));
fo.write(v.data)
def write_int4(fo, v):
c_min = np.expand_dims(-np.abs(v).max(axis = -1), -1)
c_max = np.expand_dims(np.abs(v).max(axis = -1), -1)
c_scale = c_max / 7.0
c_min = c_scale * -8.0
# c_min = np.expand_dims(-np.abs(v).max(axis = -1), -1)
# c_max = np.expand_dims(np.abs(v).max(axis = -1), -1)
# c_scale = c_max / 7.0
# c_min = c_scale * -8.0
c_min = np.expand_dims(v.min(axis = -1), -1)
c_max = np.expand_dims(v.max(axis = -1), -1)
c_scale = (c_max - c_min) / 15.0
c_zero = np.round(0.0 - c_min / c_scale)
c_zero = c_zero.clip(0, 15)
c_min = -c_scale * c_zero
v = (v - c_min) / c_scale
v = (v + 0.5).astype(np.int8).clip(0, 15).astype(np.uint8)
v = v[:, 0::2] * 16 + v[:, 1::2]
fo.write(struct.pack('i', 8))
fo.write(struct.pack('i', 0))
for i in range(c_min.shape[0]):
fo.write(struct.pack('f', c_min[i][0]))
fo.write(struct.pack('f', c_max[i][0]))
fo.write(struct.pack('f', c_min[i][0]));
fo.write(struct.pack('f', c_max[i][0]));
fo.write(v.data)
def tofile(exportPath,
......@@ -91,8 +99,14 @@ def tofile(exportPath,
# Baichuan 2代
modelInfo["use_alibi"] = "1"
modelInfo["pre_prompt"] = ""
modelInfo["user_role"] = ("<FLM_FIX_TOKEN_" + str(model.generation_config.user_token_id) + "> ") if hasattr(model.generation_config, "user_token_id") else ""
modelInfo["bot_role"] = ("<FLM_FIX_TOKEN_" + str(model.generation_config.assistant_token_id) + ">") if hasattr(model.generation_config, "assistant_token_id") else ""
modelInfo["user_role"] = ("<FLM_FIX_TOKEN_" + str(model.generation_config.user_token_id) + ">") if hasattr(model.generation_config, "user_token_id") else "";
modelInfo["bot_role"] = ("<FLM_FIX_TOKEN_" + str(model.generation_config.assistant_token_id) + ">") if hasattr(model.generation_config, "assistant_token_id") else "";
modelInfo["history_sep"] = ""
if (modelInfo["model_type"] == "baichuan" and modelInfo["vocab_size"] == 125696):
# Baichuan 2代 7B
modelInfo["pre_prompt"] = ""
modelInfo["user_role"] = ("<FLM_FIX_TOKEN_" + str(model.generation_config.user_token_id) + ">") if hasattr(model.generation_config, "user_token_id") else "";
modelInfo["bot_role"] = ("<FLM_FIX_TOKEN_" + str(model.generation_config.assistant_token_id) + ">") if hasattr(model.generation_config, "assistant_token_id") else "";
modelInfo["history_sep"] = ""
if modelInfo["model_type"] == "qwen":
if modelInfo["chat_format"] == "chatml":
......@@ -140,10 +154,10 @@ def tofile(exportPath,
for v in vocab.keys():
if (modelInfo['model_type'] == "qwen"):
s = v
elif (modelInfo["model_type"] == "moss"):
s = [(ord(c) if c not in tokenizer.byte_decoder else tokenizer.byte_decoder[c]) for c in v]
else:
s = v.encode()
if (modelInfo["model_type"] == "moss"):
s = [(ord(c) if c not in tokenizer.byte_decoder else tokenizer.byte_decoder[c]) for c in v]
fo.write(struct.pack('i', len(s)))
for c in s:
fo.write(struct.pack('i', c))
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment